MARKETING
RESEARCH
This page is intentionally left blank
MARKETING
RESEARCH
ELEVENTH EDITION
DAVID A. AAKER
E. T. Grether Professor of Marketing and Public Policy
University of California, Berkeley
V. KUMAR
Richard and Susan Lenny Distinguished Chair Professor
in Marketing,
Director of the Marketing Ph.D. Program, and
Executive Director, Center for Excellence in Brand and Customer
Management
Georgia State University
ROBERT P. LEONE
J. Vaughn and Evelyne H. Wilson Chair and
Professor of Marketing
Neeley School of Business, Texas Christian University
GEORGE S. DAY
Geoffrey T. Boisi Professor of Marketing, and
Director of the Huntsman Center
for Global Competition and Innovation
Wharton School, University of Pennsylvania
Dedicated with love
To my wife, Kay
To my parents, Patta and Viswanathan
my family—Aparna, Anita and Rohan, and Prita
my sister Shanti, and her family, Prasad, Amritha and Deepa,
my uncle, Kannan
and, my in-laws, Dr. Lalitha and Ramamurthy
To my wife, Teresa
and my daughters—Amy and Casey
To my wife, Alice
VICE PRESIDENT AND EXECUTIVE PUBLISHER George Hoffman
SENIOR ACQUISITIONS EDITOR Franny Kelly
PROJECT EDITOR Brian Baker
ASSOCIATE DIRECTOR OF MARKETING Amy Scholz
MARKETING MANAGER Kelly Simmons
MARKETING ASSISTANT Marissa Carroll
PRODUCT DESIGNER Allison Morris
MEDIA SPECIALIST Ethan Bernard
ASSOCIATE PRODUCTION MANAGER Joyce Poh
PHOTO DEPARTMENT MANAGER Hilary Newman
COVER DESIGNER Wendy Lai
COVER PHOTO Datacraft Co. Ltd./Getty Images, Inc.
This book was set in 10/12 Palatino Roman by Aptara Corporation and printed and bound by Courier Westford.
This book is printed on acid free paper.
Copyright © 2013, 2010, 2007 John Wiley & Sons, Inc. All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording,
scanning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to
the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, website www.copyright.com. Requests to
the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street,
Hoboken, NJ 07030-5774, (201)748-6011, fax (201)748-6008, website http://www.wiley.com/go/permissions.
To order books or for customer service, please call 1-800-CALL WILEY (225-5945).
Library of Congress Cataloging-in-Publication Data
Marketing research/David A. Aaker . . . [et al.]. — 11th ed.
p. cm.
Includes index.
ISBN 978-1-118-15663-6 (cloth) 1. Marketing research. I. Aaker, David A.
HF5415.2.A14 2013
658.8’3—dc23
2012027801
Printed in the United States of America
10
9
8
7
6
5
4
3
2
1
PREFACE
T
he realm of marketing research truly flourishes as more data becomes available to marketers. Using data, companies continue to glean important insights that place them on
the road to marketing and financial success. Ever since companies started to collect primary data through consumer surveys to help them plan marketing activities, the field
of marketing research with respect to the data requirements has come a long way. The nature
and type of data being collected now also includes scanner data, and more recently social media
data. The survey and scanner data techniques have empowered companies to collect a wide
range of customer information pertaining to attitudes, behaviors, and demographics. Ably supported by the growth of data analysis software (e.g., SAS, SPSS, R, etc.), and data storage and
retrieval techniques (ranging from small databases to huge data farms), the area of marketing
research has grown by leaps and bounds over the past decades. This growth has been accentuated by the declining data storage costs and the growing expanse of the Internet medium. These
developments have enabled companies to invest in data now more than ever. These forces, coupled with the technology trends and the emergence of a global marketplace, require managers
to take cognizance of the important role of marketing intelligence. Further, the customer level
interactions on the Internet and the emergence of social networking as an important medium of
marketing adds a new dimension to traditional marketing research intelligence. This new edition of Marketing Research brings to the forefront the relevance of marketing intelligence and the
power of the Internet in marketing research applications by tracking recent developments in
social media pertaining to the popularity of social coupons and the prevalence of social media
marketing.
If we can compare marketing to a long train with multiple compartments, then marketing
research would justly claim the dual roles of the engine that powers the train and the links that
connect the individual compartments to form a cohesive functional unit. In other words, marketing research is pervasive—the brain and the brawn of any marketing organization. Having said
this, we realize that marketing research is a complex subject and therefore has to be introduced
to the student one compartment at a time before the entire train can be visualized. We also realize the danger in this approach. The student can get overly excited or, even worse, overwhelmed
by the individual units so that he or she fails to see the proverbial “big picture”—the overarching framework, the subtle but essential interactions between units, and the ultimate purpose,
namely, how marketing research can help organizations achieve their goals. Hence, this text takes
a “macro-micro-macro” approach toward communicating the intricacies of marketing research
and its usefulness to the marketing organization.
This revised edition, while maintaining the strengths of previous editions, focuses on the recent
trends in marketing intelligence and explores how the popularity and spread of social media influences continue to have an enormous impact on marketing research applications. Topics of less interest and relevance to the practice of marketing and marketing research have been eliminated.
New topics have been added to position this book as the necessary tool for successful marketing
management in the twenty-first century.
This edition, while building on the online marketing research content, tracks recent developments in the area of relationship marketing. Specific importance has been given to the concepts of Customer Lifetime Value, Mobile Marketing, Social Media Marketing and Experiential
Marketing. Together, these address the advancement of marketing research to the forefront of
business strategies that involve the increased effort to win customers’ share of purchases.
v
PREFACE
We begin with a macro-level treatment of what marketing research is, where it fits within an
organization, and how it helps in managerial decision making. Here, we also discuss the marketing research industry, with a brief treatment of both suppliers and users.
The body of the text takes a micro-level approach, detailing each and every step of the marketing research process. In describing the marketing research process, a decision-oriented perspective
has been adopted to help students, who are future managers and researchers, make better decisions.
Detailed discussions of the process, with numerous examples from the industry, characterize this
micro phase.
Finally, we wrap up with a macro-level treatment of the applications of marketing research.
Here we address the traditional 4P research, as well as contemporary issues such as brand equity,
customer satisfaction research, and emerging issues that continue to fascinate marketers, such as
e-commerce, direct marketing, database marketing, and relationship marketing, while taking care
to incorporate some of the latest research and developments in these fields.
OBJECTIVES OF THIS TEXT
Our overall objectives in writing this text continue to be:
1. To emphasize the role of marketing intelligence in today’s world and to focus on the techniques and steps that show how a company can gather marketing intelligence and to examine its role in strategic decision making.
2. To communicate in an interesting and informative manner the essence of marketing research
to “future managers” and “future researchers.” Both groups need to know when marketing
research can and should be used, what research alternatives exist, how to recognize effective
and ineffective research, and how to interpret and apply the results.
3. To illustrate the usefulness of the Internet, online marketing research, and other advances in
technology and their applications.
4. To emphasize the current developments in marketing research, such as the distinction between
domestic and international market research, the emphasis on customer value management,
and the influence of social media.
5. To use examples, applications, and illustrations throughout the book, in an effort to tie the
material to the real world and thus provide interest and better understanding to the student.
6. To discuss the fastest growing applications of marketing research, e-commerce, direct marketing,
database marketing research, customer relationship management, and social networking and
their impact on businesses.
7. To provide a clear and comprehensive treatment of modern data analysis topics. Each chapter includes simple numerical examples to help students get a hands-on feel for the material.
8. To provide thorough coverage of the most advanced and current marketing research methodologies, pointing out their limitations as well as their potential for enhancing research results.
NEW TO THIS EDITION
Consistent with these objectives, the eleventh edition has undergone some critical changes. The
more prominent of these include:
vi
1. The chapter on Marketing Research on the Internet has undergone a major revision with the
addition of new material on topics such as Google Analytics, trends in on-line advertising
spending, and changes in the international arena.
2. The chapter on Survey Methods has been revised to include the large number of selfenumeration survey methods that are now employed by marketing researchers including
web-based surveys and email surveys that have replaced the more traditional mail surveys
that have become less effective for many target respondents.
PREFACE
3. The section on Special Topics in Data Analysis has undergone a major overhaul with
increased focus provided to important univariate and multivariate data analysis techniques
that are widely used in business applications.
4. The section on Applications of Marketing Intelligence has been revised and re-arranged along
the lines of measures, metrics, and strategies that help companies in implementing the concepts
of marketing research.
5. The chapter on Marketing-Mix Measures focuses on more recent advances in this area such as
the development and use of dashboards, measuring advertising effectiveness across the new
platform of choices available from TV to websites to social media opportunities.
6. The chapter on Brand and Customer Metrics reviews topics such as how to maximize
customer profitability through customer selection, retention, referral programs and word
of mouth.
7. The chapter on New Age Strategies contains material on topics such as mobile marketing, the
use of social media in marketing (e.g. blogs, social networks, content communities, etc.).
8. Statistics and other trend information have been updated given new insights offered by the
2010 US Census.
9. New cases and study questions have been added. These relate to current problems facing marketing researchers, such as how to reach subjects by phone given the large number of people
who no longer have land lines, the use of omnibus survey panels, focus group design, and
web-based marketing research.
10. The text has also benefitted from additions and updates to new and emerging topics of
interest and methods of practice in marketing research, such as maximizing customer
profitability, social media marketing, and word-of-mouth marketing.
11. The web address www.wiley.com/college/aaker can be used for accessing information
pertaining to the textbook and its contents.
FEATURES OF THE BOOK
1. The text communicates in an interesting and informative manner the essence of marketing
research to future managers and future researchers. Both groups need to know when marketing research can and should be used, what research alternatives exist, how to recognize
effective and ineffective research, and how to interpret and apply the results.
2. The Cases and Part Cases are positioned appropriately at the end of chapters and text parts
to stimulate interest, add realism to the marketing research curriculum, and help develop
decision-making skills. These cases cover a wide range of products and organizations.
3. The chapter on A Decision-Making Perspective on Marketing Intelligence attempts to enlighten
the readers of the importance and means of accessing data from multiple sources and delivering to end-users for analysis.
4. The chapter on Marketing Research on the Internet links the reader with the world of marketing research to keep abreast with the emerging trends and changes in the marketplace.
5. Discussion of the international element of marketing research has been continued. Particularly, an effort has been made to provide a clear distinction between the domestic and international marketing research processes and prepare the users of this text to face the challenges
of multinational research.
6. The chapter on New Age Strategies reiterates the fact that understanding of the customer is the
key to marketing success. The chapter focuses on the tools provided by marketing intelligence
for better knowledge of customer profiles. The chapter also focuses on e-commerce, database
marketing, relationship marketing and social networking. The growth in e-commerce is phenomenal and the firm’s ability to identify individuals and market to them is an important
task. Database marketing is currently one of the most important tools for businesses facing
the challenges of the twenty-first century. As firms shift their resources more towards
targeted marketing, the discussion in this chapter becomes more valuable.
vii
PREFACE
7. Marketing Research in Action sections in various chapters have been updated to focus on
the real world applications of Marketing Research.
8. Each chapter also includes Learning Objectives, a Chapter Summary, and Questions and
Problems.
ORGANIZATION OF THE TEXT
The book is organized to reflect the “macro-micro-macro” approach towards imparting marketing research training to the student. The text consists of five parts. Parts I and V deal with the
“macro” aspects of marketing research; Parts II, III, and IV deal with the micro aspects.
Part I, consisting of four chapters, deals with the nature and scope of marketing intelligence
and marketing research. Here, the overall framework of marketing research is presented,
and where and how marketing research fits in with the other aspects of marketing is
explained. The nature of the research industry and suppliers is also discussed here.
Part II, consisting of Chapters 5 through 15, deals extensively with the various aspects of data
collection. This part is further divided into four sections, one section devoted to each of the
three fundamental types of marketing research: exploratory, descriptive, and causal. The
final section addresses the issue of sampling.
Part III, consisting of three chapters, discusses the fundamental aspects and techniques
in data analysis. These include basic analysis issues such as data editing, coding, and
simple techniques such as hypothesis testing, chi-square analysis, and the analysis of
variance.
Part IV is devoted exclusively to advanced and more sophisticated data analysis techniques
such as correlation and regression analysis, discriminant analysis, factor analysis, cluster analysis, conjoint analysis, and multidimensional scaling. This part consists of four
chapters, the last chapter dealing with the aspects of presenting the research results.
In Part V, the student is exposed to the measures, metrics, and strategies of marketing intelligence. This section has three chapters and provides the student with a comprehensive
picture of marketing research, highlighting how the marketing-mix measures, the brand
and customer metrics, and the new age strategies fit into the application of marketing
intelligence.
SUPPLEMENTS TO THE TEXT
■ All relevant information pertaining to the textbook and its contents can be accessed from this
website—www.wiley.com/college/aaker. This website will periodically update information
that is relevant for keeping the text up to date.
■ An Instructor’s Manual will accompany this text. This manual provides solutions to endof-chapter Questions and Problems, and discusses all the test cases in greater detail. Exam
questions are arranged by chapter, and include multiple-choice and true/false questions. An
example of a course syllabus is presented, and many suggestions for the organization of the
course are provided. This resource can be found on the Instructor Companion site.
■ A test bank has been added to the website, which provides sample test questions in each
chapter.
■ PowerPoint presentations of selected material from each chapter are available.
■ Data sets that contain data for end-of-chapter cases and statistical analysis are available.
SPSS®-interpreted examples are also available.
viii
PREFACE
■ The website also contains detailed information on the mini-project designed for students mentioned at the end of the eighth chapter.
■ End-of-chapter questions are also available on the website.
■ New Marketing Research Video Series: Brand new interview-style video clips of top marketing research companies. Each video, six to eight minutes in length, presents interviews with
key personnel to discuss how they apply the major concepts of marketing research to their
business. The Wiley Marketing Research Video Series can be accessed on the Instructor and
Student Companion websites.
ACKNOWLEDGMENTS
Many debts have been accumulated over the years during which ten editions of this book have
taken shape. We are especially grateful to our students, who gave us feedback from the consumer’s
perspective and whose field research projects provided many of the illustrations and problems; to
our colleagues, who stimulated us and brought new ideas and approaches to our attention; and to
our clients, who gave us many opportunities to put ideas into practice and thus broaden our understanding of marketing research as it is currently practiced. It has been a continuing pleasure to associate with a class publisher, John Wiley & Sons, and to work with a number of Wiley editors over
the years—Rich Esposito, John Woods, Tim Kent, Ellen Ford, Brent Gordon, as well as Jeff Marshal
and Cindy Rhoads, Jayme Heffler and Jennifer Conklin, and most recently, Franny Kelly.
A host of helpful and insightful reviews were received from Miodrag Lovric from University of Belgrade, Curt J. Dommeyer from California State University, S. Chan Choi of Rutgers
Business School, Wayne A. Roberts, Jr. of Southern Utah University, L. J. Shrum of the University of Texas at San Antonio, Philip M. Hurdle of Elmira College, Michael R. Hyman of New
Mexico State University, and Eric Vincent Carter of Oakland University.
We also would like to express thanks to Bharath Rajan, A. J. Radcliffe and Sarang Sunder for
their efforts towards the production of the eleventh edition.
Finally, we wish to express our sincere appreciation to our families and friends for their constant support, encouragement, and sacrifices during the creation of this book.
David A. Aaker
V. Kumar
Robert P. Leone
George S. Day
ix
This page is intentionally left blank
BRIEF CONTENTS
PART I
The Nature and Scope of Marketing Research
Chapter 1 A Decision-Making Perspective on
Marketing Intelligence, 1
Chapter 2 Marketing Research in Practice, 29
Chapter 3 The Marketing Research Process, 45
Chapter 4 Research Design and
Implementation, 72
PART III
Data Analysis
Chapter 16 Fundamentals of Data
Analysis, 403
PART II
Data Collection
PART IV
Special Topics in Data Analysis
Chapter 19 Correlation Analysis and
Regression Analysis, 476
SECTION A
Secondary and Exploratory Research
Chapter 5 Secondary Sources of Marketing
Data, 102
Chapter 6 Standardized Sources of
Marketing Data, 126
Chapter 7 Marketing Research on the
Internet, 152
Chapter 8 Information Collection:
Qualitative and Observational Methods, 177
SECTION B
Descriptive Research
Chapter 9 Information from Respondents:
Issues in Data Collection, 215
Chapter 10 Information from Respondents:
Survey Methods, 231
Chapter 11 Attitude Measurement, 260
Chapter 12 Designing the Questionnaire, 288
Chapter 17 Hypothesis Testing: Basic
Concepts and Tests of Associations, 422
Chapter 18 Hypothesis Testing: Means and
Proportions, 440
Chapter 20 Discriminant, Factor, and Cluster
Analysis, 506
Chapter 21 Multidimensional Scaling and
Conjoint Analysis, 548
Chapter 22 Presenting the Results, 573
PART V
Applications of Marketing Intelligence
Chapter 23 Marketing-Mix Measures, 591
Chapter 24 Brand and Customer
Metrics, 635
Chapter 25 New Age Strategies, 655
Appendix: Tables, 694
A-1. Standard Normal Probabilities, 694
A-2. X2 Critical Points, 695
A-3. F Critical Points, 697
A-4. t Critical Points, 702
SECTION C
Causal Research
Chapter 13 Experimentation, 323
A-5. Procedures for Conducting
Univariate and Multivariate Analysis
in SPSS, 703
A-6. Output of Select Tables in SPSS, 707
SECTION D
Sampling
Chapter 14 Sampling Fundamentals, 353
Chapter 15 Sample Size and Statistical
Theory, 380
Glossary, 718
Index, 729
xi
CONTENTS
PART I
THE NATURE AND SCOPE OF MARKETING
RESEARCH
CHAPTER 1 A Decision-Making Perspective
on Marketing Intelligence, 1
End Notes, 44
Appendix: Marketing Research Jobs, 44
Learning Objectives, 1
An Overview of Business Intelligence, 1
CHAPTER 3 The Marketing Research Process, 45
Introduction to Marketing Intelligence, 3
Learning Objectives, 45
Overview of the Marketing Research Process, 45
The Preliminary Stages of the Marketing Research
Process, 47
Planning a New HMO, 56
The International Marketing Research Process, 59
Summary, 62
Questions and Problems, 62
End Notes, 63
Appendix: The Value of Research Information Using
Bayesian Decision Theory, 63
Case 3-1: A VideOcart Test for Bestway Stores, 66
Case 3-2: Sperry/MacLennan Architects and
Planners, 67
Case 3-3: Philip Morris Enters Turkey, 70
Marketing Research, 6
Role of Marketing Research in Managerial Decision
Making, 9
Factors That Influence Marketing Research
Decisions, 17
Use of Marketing Research, 18
Ethics in Marketing Research, 19
The Respondent’s Ethics and Rights, 20
International Marketing Research, 22
Summary, 24
Questions and Problems, 24
End Notes, 24
Case 1-1: Preteen Market—The Right Place to Be In
for Cell Phone Providers?, 26
Case 1-2: Best Buy on a Segmentation Spree, 26
Case 1-3: Ethical Dilemmas in Marketing Research, 28
CHAPTER 2 Marketing Research in Practice, 29
Learning Objectives, 29
Information System, Decision Support Systems,
and Marketing Research, 30
Marketing Decision Support Systems, 32
Suppliers of Information, 36
Criteria for Selecting External Suppliers, 41
The International Marketing Research Industry, 41
Career Opportunities in Marketing Research, 42
Summary, 43
Questions and Problems, 44
xii
CHAPTER 4 Research Design and
Implementation, 72
Learning Objectives, 72
Research Approach, 74
Research Tactics and Implementation, 80
Budgeting and Scheduling the Research Project, 82
Research Proposal, 83
Designing International Marketing Research, 86
Issues in International Research Design, 87
Summary, 89
Questions and Problems, 90
End Notes, 90
Appendix: Errors in Research Design, 90
Case 4-1: Reynolds Tobacco’s Slide-Box
Cigarettes, 92
Case 4-2: California Foods Corporation, 96
PART II
Case for Part I, 99
Case I-1: Clover Valley Dairy Company, 99
DATA COLLECTION
SECTION A: SECONDARY AND
EXPLORATORY RESEARCH
CHAPTER 5 Secondary Sources of Marketing
Data, 102
Learning Objectives, 102
Secondary Data, 102
Uses of Secondary Data, 103
Benefits of Secondary Data, 104
Limitations of Secondary Data, 104
Internal Sources of Secondary Data, 105
External Sources of Secondary Data, 107
Census Data, 112
North American Industry Classification
System, 114
Appraising Secondary Sources, 115
Applications of Secondary Data, 116
Sources of Secondary Data for International
Marketing Research, 119
Problems Associated with Secondary Data in
International Research, 120
Applications of Secondary Data in International
Research, 120
Summary, 121
Questions and Problems, 122
End Notes, 122
Case 5-1: Barkley Foods, 123
Case 5-2: Dell in Latin America?, 124
Case 5-3: Eddie Bauer: Strategize with Secondary
Marketing Data, 124
CHAPTER 6 Standardized Sources of Marketing
Data, 126
Learning Objectives, 126
Retail Store Audits, 127
Consumer Purchase Panels, 127
Scanner Services and Single-Source
Systems, 131
Media-Related Standardized
Sources, 137
Applications of Standardized Sources
of Data, 140
Summary, 141
Questions and Problems, 141
End Notes, 142
Case 6-1: Promotion of Rocket Soups, 143
Case 6-2: Kerry Gold Products, Ltd., 144
Case 6-3: Paradise Foods, 144
CHAPTER 7 Marketing Research
on the Internet, 152
Learning Objectives, 152
Current Trends in Web Usage, 152
WWW Information for Marketing
Decisions, 154
The Internet and Marketing Research Today, 158
The Internet and Marketing Research Developments, 168
Issues and Concerns, 171
Summary, 174
Questions and Problems, 174
End Notes, 175
Case 7-1: Caring Children’s Hospital, 176
CHAPTER 8 Information Collection: Qualitative
and Observational Methods, 177
Learning Objectives, 177
Need for Qualitative Research, 177
Qualitative Research Methods, 178
Observational Methods, 197
Recent Applications of Qualitative and
Observational Methods, 202
Summary, 204
Questions and Problems, 204
End Notes, 205
Appendix: Myths of Qualitative Research: It’s
Conventional, But Is It Wisdom?, 207
Case 8-1: Mountain Bell Telephone Company, 209
Case 8-2: U.S. Department of Energy (A), 211
Case 8-3: Acura, 211
Case 8-4: Exploratory Research on the Lleyton Hewitt
National Visa Card, 212
Case 8-5: Hamilton Beach Conducts Primary Research
in Mexico and Europe, 213
Mini Project, 214
xiii
SECTION B: DESCRIPTIVE RESEARCH
CHAPTER 9 Information from Respondents:
Issues in Data Collection, 215
Learning Objectives, 215
Information from Surveys, 215
Sources of Survey Error, 216
Methods of Data Collection, 221
Factors Affecting the Choice of a Survey Method, 221
Ethical Issues in Data Collection, 225
Summary, 227
Questions and Problems, 227
End Notes, 228
Case 9-1: Essex Markets, 229
Case 9-2: More Ethical Dilemmas in Marketing
Research, 229
CHAPTER 10 Information from Respondents:
Survey Methods, 231
Learning Objectives, 231
Collecting Data, 231
Surveys in the International Context, 251
Summary, 254
Questions and Problems, 255
End Notes, 256
Case 10-1: Roland Development Corp., 257
Questionnaire Design for International Research, 306
Summary, 308
Questions and Problems, 309
End Notes, 312
Case 12-1: Wine Horizons, 313
Case 12-2: Smith’s Clothing (A), 315
Case 12-3: Compact Lemon, 316
Case 12-4: Project DATA: An Urban Transportation
Study, 316
SECTION C: CAUSAL RESEARCH
CHAPTER 13 Experimentation, 323
Learning Objectives, 323
Descriptive versus Experimental Research, 324
What Constitutes Causality?, 325
Laboratory and Field Experiments, 327
Threats to Experimental Validity, 327
Issues in Experimental Research, 330
Types of Experimental Designs, 332
Guidelines for Conducting Experimental Research, 345
Limitations of Experiments, 346
Summary, 348
Questions and Problems, 349
End Notes, 350
Case 13-1: Evaluating Experimental Designs, 350
Case 13-2: Barrie Food Corporation, 351
CHAPTER 11 Attitude Measurement, 260
Learning Objectives, 260
What Are Attitudes?, 261
The Concept of Measurement and Scaling, 262
Types of Attitude Rating Scales, 265
General Guidelines for Developing a Multiple-Item
Scale, 276
Interpreting Attitude Scales, 277
Choosing an Attitudinal Scale, 278
Accuracy of Attitude Measurements, 280
Scales in Cross-National Research, 283
Summary, 283
Questions and Problems, 284
End Notes, 285
Case 11-1: National Kitchens, 286
SECTION D: SAMPLING
CHAPTER 14 Sampling Fundamentals, 353
Learning Objectives, 353
Sample or Census, 353
Sampling Process, 354
Probability Sampling, 360
Nonprobability Sampling, 366
Shopping Center Sampling, 370
Sampling in the International Context, 372
Summary, 374
Questions and Problems, 375
End Notes, 376
Case 14-1: Exercises in Sample Design, 377
Case 14-2: Talbot Razor Products Company, 378
CHAPTER 12 Designing the Questionnaire, 288
Learning Objectives, 288
Planning What to Measure, 289
Formatting the Question, 291
Question Wording: A Problem of Communication, 296
Sequence and Layout Decisions, 300
xiv
CHAPTER 15 Sample Size and Statistical
Theory, 380
Learning Objectives, 380
Determining the Sample Size: Ad Hoc Methods, 380
Population Characteristics/Parameters, 382
Sample Characteristics/Statistics, 383
Sample Reliability, 384
Interval Estimation, 386
Sample Size Question, 387
Proportions, 389
Coefficient of Variation, 390
Stratified Sampling, 392
Multistage Design, 393
Sequential Sampling, 393
Part III
Learning Objectives, 403
Preparing the Data for Analysis, 404
Strategy for Data Analysis, 408
Cross-Tabulations, 411
Factors Influencing the Choice of Statistical
Technique, 413
An Overview of Statistical Techniques, 415
Summary, 419
Questions and Problems, 419
End Notes, 420
Appendix: Choice of a Statistical Package, 421
CHAPTER 17 Hypothesis Testing: Basic
Concepts and Tests of Associations, 422
Learning Objectives, 422
The Logic of Hypothesis Testing, 423
Steps in Hypothesis Testing, 424
Basic Concepts of Hypothesis Testing, 424
Cross-Tabulation and Chi-Square, 428
Summary, 436
Questions and Problems, 436
End Notes, 437
Case 17-1: Medical Systems Associates: Measuring
Patient Satisfaction, 437
Case II-1: Currency Concepts International, 397
Appendix: Interview Guide for International
Travelers (U.S. Citizens), 402
CHAPTER 18 Hypothesis Testing: Means and
Proportions, 440
Learning Objectives, 440
Commonly Used Hypothesis Tests in Marketing
Research, 440
Relationship between Confidence Interval and
Hypothesis Testing, 443
Analysis of Variance (ANOVA), 445
The Probability-Values (p-Values) Approach to
Hypothesis Testing, 457
Effect of Sample Size and Interpretation
of Test Results, 458
Summary, 459
Questions and Problems, 459
End Notes, 460
Case 18-1: American Conservatory Theater, 461
Case 18-2: Apple Appliance Stores, 463
Cases for Part III, 464
Case III-1: The Vancouver Symphony Orchestra, 464
Case III-2: Popular Pizzas: Identifying Consumer
Preferences, 471
SPECIAL TOPICS IN DATA ANALYSIS
CHAPTER 19 Correlation Analysis and
Regression Analysis, 476
Learning Objectives, 476
Correlation Analysis, 477
Regression Analysis, 481
Summary, 499
Case for Part II, 397
DATA ANALYSIS
CHAPTER 16 Fundamentals of Data
Analysis, 403
PART IV
Summary, 394
Questions and Problems, 394
End Notes, 396
Questions and Problems, 500
End Notes, 502
Case 19-1: The Seafood Grotto, 502
Case 19-2: Ajax Advertising
Agency, 503
Case 19-3: Election Research, Inc., 504
xv
CHAPTER 20 Discriminant, Factor, and Cluster
Analysis, 506
Learning Objectives, 506
Discriminant Analysis, 506
Multiple Discriminant Analysis, 515
Summary of Discriminant Analysis, 518
Factor and Cluster Analysis, 518
Factor Analysis, 519
Summary of Factor Analysis, 528
Cluster Analysis, 530
Summary of Cluster Analysis, 540
Questions and Problems, 540
End Notes, 541
Case 20-1: Southwest Utility, 542
Case 20-2: Store Image Study, 543
Case 20-3: Behavioral Research, 545
CHAPTER 21 Multidimensional Scaling
and Conjoint Analysis, 548
Learning Objectives, 548
Multidimensional Scaling, 548
Attribute-Based Approaches, 548
Summary of MDS, 557
PART V
CHAPTER 22 Presenting the Results, 573
Learning Objectives, 573
Guidelines for Successful Presentations, 573
Written Report, 579
Oral Presentation, 583
Relationship with the Client, 586
Summary, 586
Questions and Problems, 586
End Notes, 586
Cases for Part IV, 587
Case IV-1: Smith’s Clothing (B), 587
Case IV-2: Newfood, 587
APPLICATIONS OF MARKETING INTELLIGENCE
CHAPTER 23 Marketing-Mix Measures, 591
Learning Objectives, 591
New-Product Research, 592
Pricing Research, 602
Distribution Research, 603
Promotion Research, 610
Total Quality Management, 620
Information Requirements for Total Quality
Management, 621
Summary, 627
Questions and Problems, 628
End Notes, 629
Case 23-1: Brown Microwave, 631
Case 23-2: National Chemical
Corporation, 632
Case 23-3: U.S. Department of Energy (B), 632
Case 23-4: Hokey Pokey is Born in India, 633
xvi
Conjoint Analysis, 558
Summary of Conjoint Analysis, 565
Questions and Problems, 565
End Notes, 566
Case 21-1: Nester’s Foods, 567
Case 21-2: Pepsi-Cola, 568
Case 21-3: The Electric Truck Case, 571
Case 21-4: Fargo Instruments, 572
CHAPTER 24 Brand and Customer Metrics, 635
Learning Objectives, 635
Competitive Advantage, 635
Brand Equity, 637
Customer Satisfaction, 643
Customer Satisfaction Measurement Process, 645
Contemporary Applications of Marketing
Intelligence, 649
Maximizing Customer Profitability, 650
Summary, 653
Questions and Problems, 653
End Notes, 654
CHAPTER 25 New Age Strategies, 655
Learning Objectives, 655
Database Marketing, 656
E-Commerce, 665
Mobile Marketing, 669
Social Marketing, 672
Experiential Marketing, 676
Relationship Marketing, 677
Recent Developments in Relationship Marketing, 678
Word-of-Mouth Marketing, 683
Customer Intelligence, 685
Summary, 689
Questions and Problems, 689
End Notes, 690
Appendix: Forecasting, 690
End Notes for Appendix, 693
Appendix: Tables, 694
A-1. Standard Normal, Cumulative Probability in
Right-Hand Tail for Positive Values of z; Areas Are
Formed by Symmetry, 694
A-2. X2 Critical Points, 695
A-3. F Critical Points, 697
A-4. t Cut-off Points for the Student’s
t-Distribution, 702
A-5. Procedures for Conducting Univariate and
Multivariate Analysis in SPSS, 703
A-6. Output of Select Tables in SPSS, 707
Glossary, 718
Index, 729
xvii
This page is intentionally left blank
A DECISION-MAKING PERSPECTIVE
ON MARKETING INTELLIGENCE
1
Learning Objectives
■ Understand the concept of business intelligence.
■ Understand the need and use of marketing intelligence in an organization.
■ Comprehend how marketing intelligence fits in the bigger scheme of the marketing
environment.
■ Explain the role of marketing intelligence in decision making.
■ Discuss the factors that affect marketing intelligence decisions.
■ Understand the implication of ethical issues in gathering marketing intelligence.
AN OVERVIEW OF BUSINESS INTELLIGENCE
As they manage their businesses and compete in a global market, decision makers face many questions every day. Is the business healthy? Who are my best customers? What supplier should I choose?
Where are we quarter-to-date? Do we deliver products consistently on time? Do we have the right mix
of people?
Effectively managing the performance of the business means knowing what questions to ask
and having the facts readily at hand to answer them. This is what business intelligence (BI) delivers. BI, at its core, is the ability to access data from multiple sources within an enterprise and
deliver it to business users for analysis.1 BI bridges the gap between disparate operational systems and data-hungry end users. It connects people to their business. It creates an information
environment that makes it easy for people to get the reports they need in the context of their
day-to-day activities. It provides an accessible means of analyzing the business and getting to
the bottom of what’s behind trends and anomalies. And, it offers a reliable barometer of how
well the business is performing. The following examples give a flavor of the possibilities BI offers.
Erlanger (Ky.) Police Department2
In a novel attempt at fighting crime, the Erlanger, Ky., police department has developed a new
portal that helps its officers to track and solve crimes. Developed in September 2008, the portal
combines concepts from business intelligence, searching and mapping. This Web-based BI portal allows patrol officers to enter data—or even pieces of data such as a few numbers from a
license plate—into a simple search interface and retrieve information from their own databases
and those of neighboring towns. This solution was developed to address the long standing issue
of an unified communications network bridging the fire and police departments. Specifically, the
Erlanger Police Department could not tap into the records management systems of 19 separate
government agencies in order to search and analyze information about suspects, reported incidents, arrests and crimes.
1
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
In order to develop this software, the department bought a readily available software that
has geo-mapping and BI features and added an open-source search, which compiles structured
and unstructured data references into an index that can easily be searched by an appliance. This
final software categorized search results to give users a better context and relevancy for search
queries. This Web-based software combined real-time crime data from multiple agencies with
crime records and incident reports stretching back five years to link information about suspects,
incidents and arrests.
This software was able to completely search the records management systems and, create maps
and information that helped officers perform their duties. This software enabled officers to see over
the last 24 hours where calls have occurred and compare crime stats against last year’s data and
this year’s data.
The department believes that the software’s success will be judged in part by whether the
department can continue to handle the 5 percent to 10 percent annual increase in calls to the communications center while maintaining its recent 1 percent to 2 percent annual rate of increase in
crimes solved—all without having to add more personnel.
Mozilla Firefox3
Mozilla Labs, the research wing of Mozilla Corp., is planning observational studies aimed at 1 percent of Firefox users. The study, which will have users volunteering, intends to look at how users
utilise the browser, and the Web in general. This type of sample set, Mozilla believes, will be a representative sample as against depending on the “early adopters.” A unique feature of such a study
is that owing to the open source format of the browser, everything participating users send has to
be, and will be, published in human-readable format.
Still in the planning stages, the study will involve launching a Firefox add-on that can help
collect browsing and usage data, and provide tools to answer feedback questions. The type of data
Mozilla wishes to collect is restricted to demographic information, such as their technical level and
geographic location. Further, they are interested in the number of times users visit a particular Website, and not the actual Website. Some of the specific questions that Mozilla wants to find an answer
for are, How do people multitask in the browser? What works and what does not? How do users
use tabs, and when do they chat in the background?
The benefits derived from the launch of Firefox 3.0, an advanced version of Firefox 2.0, seems
to led to this study initiative. A key upgrade in Firefox 3.0 is that the Back button grew in size,
compared to the earlier version. The reason for this change was from the finding that people
used the Back button much more than the Forward button. However, Mozilla believes that the
decision would have been easier if they had then had a large pool of testers, and more data on
when users hit the Back button or the Forward button.
Recent Trends in Business Intelligence
Business intelligence tools and systems play a key role in the strategic planning process of a
firm. These systems allow the firm to gather, store, access, and analyze corporate data to aid in
decision making. The science of BI is being utilized by specialists in different business fields
today. Business intelligence comprises several components—financial intelligence, marketing
intelligence, accounting intelligence, and management intelligence among others, as shown in
Figure 1-1. Financial intelligence provides consultancy, forward looking financial accounting and
procurement systems, strategic business intelligence, and support. To quote the words of an industry executive, “Financial intelligence works by empowering the CFO to make critical and successful decisions, giving the CIO current yet reliable technology and the CEO quality time to think
strategically.” Accounting and management intelligence incorporate the tools and techniques of
the respective fields. In this book, we are interested in the marketing intelligence component of
business intelligence.
2
INTRODUCTION
TO
MARKETING INTELLIGENCE
Business Intelligence
Financial
Intelligence
Management
Intelligence
Marketing
Intelligence
Accounting
Intelligence
Customer Relationship Marketing/
Database Marketing
Marketing Research
Primary Sources
Secondary Sources
Standardized Sources
Provide data on
Prospects
Customers
Competition
Products/Markets
FIGURE 1-1
Typical areas under the umbrella of business intelligence.
INTRODUCTION TO MARKETING INTELLIGENCE
According to American Marketing Association, marketing is defined as the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging offerings that
have value for customers, clients, partners, and society at large.4 The marketing concept does not
focus on one aspect of marketing, but recognizes the roles of non-marketers in the marketing
process such as customers, vendors, or external agencies who regulate marketing.
While marketing centers around satisfying customer needs, an organization would often face
the requirement of obtaining information on customer needs. Marketing intelligence helps organizations in this process and revolves around gathering customer level information that would
assist in business decision making and policy analysis.
What Is Marketing Intelligence?
Market intelligence (MI) is “the process of acquiring and analyzing information in order to understand
the market (both existing and potential customers); to determine the current and future needs and preferences, attitudes and behavior of the market; and to assess changes in the business environment that may
affect the size and nature of the market in future.”5
Marketing Intelligence has the capacity to be at the forefront in contributing to the development of a business environment through strategic research, risk and policy analysis, creditrating documentation, storage, publication, reporting, and communication of reliable, timely, and
objective business information. It incorporates information from customer analysis and industry
analysis as well as general market conditions.
3
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
In other words, marketing intelligence calls for understanding, analyzing and assessing the internal and external environment related to a company’s customers, competitors, markets, and industry
to enhance the decision-making process. This would require the integration of competitive intelligence, marketing research, market analysis, and business and financial analysis information.6
Components of Marketing Intelligence
Marketing intelligence primarily focuses on creation and management of profitable long-term
relationships with customers with an emphasis on the use of information and communication
technologies in researching, selecting, entering, and competing in global markets. As shown in
Figure 1-1, the various components of marketing intelligence can be included under two main
areas—marketing research and customer relationship marketing/database marketing.
Marketing research focuses primarily on the collection and use of information on customers
and their needs for designing marketing programs. Marketing data can be collected from primary, secondary, and standardized sources. Data analysis in marketing research uses the data on
prospects, customers, competition, products, and markets while incorporating marketing concepts, methods, and quantitative tools, to make meaningful decisions with regard to marketing
campaigns, resource allocation, and the managerial planning process. Customer relationship
management/database marketing integrates information on customer buying behavior and motivation obtained from a firm’s database and market research. This information aids the firm to
manage the offers made to its customers (both current and new products and services), in order
to acquire new customers, and satisfy and retain existing customers in a profitable manner.
Need for Marketing Intelligence
Marketing research plays a critical part in a marketing intelligence system. It aids in improving
management decision making by providing relevant, accurate, and timely information. The following examples provide a flavor of marketing research in action.
■ The Subaru Outback was introduced in April 2000. The car was positioned in-between station
wagons and compact SUVs, based on findings that the sizeable male-dominated segment
needs were not met by either station wagons or SUVs. Following one of its best years in U.S.
sales in 2006, Subaru developed an all 4-wheel-drive lineup. This positioning is a shift from
its prominent truck-based SUV designs—sales of which had begun to slow. In volatile markets,
Subaru introduced the B9 Tribeca which looked like a mid-sized SUV but was based on a car
platform, taking advantage of a sweet spot between the two segments.
■ The iPod, Apple Inc.’s brand of portable music players was introduced in 2001 with revolutionary design and technology that would leverage on the brand exclusivity of the existing
iMac computers. Despite emerging the market leader and surpassing the 100 million units
sold milestone by 2007, Apple realized that the portable music device market was maturing.
The introduction of the iPhone has been as a key to maintaining sales volumes and sustaining
Apple’s dominant position in the portable music device market. Apple’s iPhone has been able
to capitalize on the successful design of the iPod, already well received by their customer base,
and to exercise their strength in innovation.7
■ A major pharmaceutical company provides its sales personnel with up-to-date information on
its products including relevant media and news clips, professional profiles, and order history
of the clients they are about to visit.
All these examples point to the need for marketing intelligence because:
■ Producers have little direct contact with consumers (geographical distance and channel layers).
■ Channels have little knowledge about customer attitudes, preference, and changing tastes.
4
■ Need to understand competition without spying on them.
INTRODUCTION
TO
MARKETING INTELLIGENCE
■ Management goals involve sales targets and market share achievement.
■ Need to identify successful new product developments early in the process to ensure growth
and revenue maximization by finding a balance between costs and prices of products.
■ The future is uncertain, yet, businesses do need to anticipate the future. The need is to look
at least up to 5–10 years in advance. Pharmaceutical companies do it on a regular basis because
they have to plan for what happens when a particular patent expires.
Table 1-1 lists the various areas where marketing intelligence is applied.
Traditional forecasting involves extrapolating from the present. The problem with this
approach is that the further forward one goes, the greater the error—thus, the greater the risk
of a mistake. Three examples illustrate the problem of anticipating the future.
In 1943, Thomas Watson, then chairman of IBM, forecasted a world market for about five computers. In 1970, Ken Olsen, founder of Digital Equipment Corporation, said no one needed to have
a personal computer at home. In 1981, Microsoft’s founder Bill Gates said that 640 K would be
enough memory for anyone.
According to AWARE, a marketing intelligence services company, Scenario Planning is a
technique that overcomes these problems. It is a methodology that allows practitioners to prepare for the future by looking at trends in the present and mapping how these interrelate with
each other. From this study, a number of scenarios picturing possible future worlds are drafted,
along with a description of how these futures could arise. Businesses then plan around these—
with contingency plans for undesirable elements. Key change indicators are monitored and used
to gauge how events are turning out.
Companies using the technique have reported remarkable successes—the oil company Shell,
for example, predicted the Oil Crisis of the 1970s, and the changes that took place prior to and
immediately after the fall of the Soviet Union in the early 1990s.
TABLE 1-1
Domains of Marketing Intelligence
■ Product decisions
Marketing intelligence helps in making decisions on investing in new products and also in optimizing product specifications or offering to target customer groups.
■ Customer segmentation decisions
Marketing intelligence gives an opportunity for companies to study the impact of advertisements—
before and after advertisements. Based on research and audience composition, marketing intelligence gives companies the flexibility to choose advertising media. Finally, MI also gives companies
enough information for selection of target segments.
■ Brand and pricing decisions
Marketing intelligence helps decision makers in building brand equity and consciousness among
the targeted customer segments. Gathering marketing intelligence also helps capture popular perceptions and changing trends, tastes, and lifestyles.
■ Keeping stakeholders happy
Understanding stakeholder needs and responding to them keeps all stakeholders—customers,
channels, and suppliers—satisfied. A satisfied customer is a profitable customer.
■ Market estimation, competitive benchmarking, and distribution
Marketing intelligence also helps in quantitative analysis of market forces—estimating size of
the market, quantitative and behavioral analysis of the market, and forecasting future trends/needs.
Marketing intelligence also helps in market share analyses.
In short, the domains of marketing intelligence application appear vast and include identifying
marketing opportunities to build profitable businesses; developing and using marketing intelligence;
designing the marketing mix; acquiring and retaining customers; and planning, organizing, evaluating, and controlling market performance.
Judicious gathering of marketing intelligence helps companies to walk up the marketing
thinking ladder by explaining how to grasp and outperform consumer value migration and hence
to make more money by adding more value to existing brands. Outcome is simply more consumeroriented businesses.
5
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
The different functions under marketing intelligence are all interrelated in the sense that the
data collected using various techniques in marketing research are then analyzed and interpreted
to improve the efficient management of customer–firm relationships, and the knowledge and
experience created from these fields can be used to effectively manage new product development and marketing. Customer relationship marketing/database marketing will be covered
briefly in the last chapter. But we delve into the marketing research arm of marketing intelligence for the most part of our discussion in this book.
MARKETING RESEARCH
Any marketing organization should have a marketing intelligence system in place that helps
them obtain information on consumer needs and gathers marketing intelligence that will help
them understand how to satisfy these needs efficiently. Each decision a company makes poses
unique needs for information gathered through marketing research, and relevant strategies can
be developed based on the information gathered through marketing research. These strategies
help companies in designing and implementing specific marketing campaigns, for example, multicultural marketing campaigns (MCM).
For instance, a survey by the Association of National Advertisers, New York in 2008 found that
nearly 77 percent of marketers implemented multicultural marketing (MCM) initiatives, with 66 percent of marketers indicating that their company’s efforts in the MCM area have increased over the
past few years. However, only 45 percent expressed satisfaction with the results of their MCM initiatives, with 26 percent saying they were “somewhat” or “very” dissatisfied. Further, firms with
MCM offerings reported that campaigns for specific target segments have increased significantly
since 2003. Table 1-2 provides the percentage of firms with MCM offerings, by target segment.
The survey also found print to be the most favored media vehicle for reaching multicultural
audiences (65%), followed closely by TV (61%), sponsorships (54%), public relations (54%), targeted radio (53%), in-store marketing (51%) and events (51%). Online advertising placed lowest
(49%). These findings have important implications in targeting and developing products and
services for the multicultural segments. The following examples give the flavor of marketing
research in targeting new market segments.
Conquering Latino Homes8
Hispanics account for nearly 16 percent of the U.S. population and are about 50.5 million strong.
According to the Census Bureau, the demographic grew 43 percent from 2000 to 2010, making it the
fastest growing demographic in the United States. The America Online/Roper ASW U.S. Hispanic
Cyber study reports that Hispanics spend nearly 9.5 hours a week online at home and another
13.8 hours a week online at work. This translates to 16 percent more time than the Internet-using
U.S. adult population as a whole. The report further notes that, when they meet new people, 43 percent hand over an e-mail address as the sole means of staying in touch compared to 31 percent of
the total U.S. online consumer population.9 This demographic holds much promise to marketers and
advertisers in terms of revenue and has been receiving increased attention over the years.
TABLE 1-2
MCM Offerings by Target Segment
6
Target Segment
2008
2003
Hispanic American
95%
86%
African American
76%
60%
Asian American
38%
35%
Gay, Lesbian, Bisexual, Transgender
24%
NA
Source: http://www.marketingcharts.com/topics/demographics/multicultural-marketing-grows-but-frustrations-mounts-6839/
MARKETING RESEARCH
The Hispanic market contributes $600 billion of the $1.3 trillion purchasing power of the
multicultural population. Despite the popularity of the Hispanic market, advertisement spending aimed at this demographic remains weak. According to the Association of Hispanic Advertising Agencies (AHAA), the nation’s top 671 advertisers, accounting for 80 percent of the total
ad spending, spent only $3 billion on print and television advertising to Latinos in 2003; though
up from $2.3 billion in 2002, the total is $1 billion less than Procter & Gamble’s global ad budget.
The association further added that top U.S. marketers spent 5.1 percent of their total advertising
budgets on the Hispanic market in 2003, up from 4.6 percent in 2002. Data compiled by AHAA
also reveal that SABMiller spent $40.1 million on Hispanic-targeted television and print advertising last year, the most by an advertiser in the alcoholic-beverage category. They are closely
followed by their rival Anheuser-Busch with $21.5 million in spending. Mintel, a Chicago-based
market research firm, reports that 75 percent of the Hispanic population is under 45 years old—
a prime age group for beer consumption. A Minkel Survey found that 41 percent of whites consumed beer, compared with 38 percent of Hispanics and 29 percent of African-Americans. Hispanics were also found to be the most brand-loyal among beer consumers. The SABMiller unit
has entered into a $100 million contract with Univision Communications, their largest ever with
a single advertiser, and a sure sign of brewers’ interest to target Hispanics. During the first half
of the 2006 World Cup soccer matches, Miller will be the exclusive beer advertiser. The cause for
Miller’s interest is a booming Hispanic population.10
Advertisers are now making ads that depict cultural integration and generate product awareness. The pioneer in this effort is Sears, which came up with the “Solo Para Ti” (Just for You)
campaign on Spanish-speaking television. Imagine The ROI of Sears in 2050 when the Hispanic
population represents 25 percent of the total U.S. population! Marketing Research in Action 1-1
provides the initiative made by Hewlett-Packard Co. in targeting Hispanic consumers.
Even Internet service providers (ISP) are taking an interest in this segment. In October 2003,
America Online launched AOL Latino, a Spanish language version of its 9.0 Optimized online
service designed to meet the needs of Hispanics in the United States. For AOL and its advertisers,
MARKETING RESEARCH IN ACTION 1-1
HP Targets Hispanic Consumers
HP created an integrated marketing program to run during the 2006 holiday season targeting Hispanic
consumers. For the holiday season, HP focused on a set of products that can be given as holiday gifts,
including notebooks, desktops, printers and cameras. In order to get the word out, HP partnered with
500 retail outlets and used a multi-channel call to action. Through the various channels, HP invited consumers to call or go online or to approach a certain retailer for more information on the products on offer.
TV and radio ads were aired in select markets with a higher concentration of Hispanic consumers,
including Miami, Houston, and San Diego. TV commercials featuring Papa Noel (Santa Claus in Spanish)
in the “PC is Personal” campaign was used to good response. This segment also included targeting Spanish
speaking DJs with sweepstakes and other gifts. The DJs would tell listeners to visit a local retailer and win,
for example, an HP camera or printer, which drove traffic to the company’s retail partners. The Web
component of the promotion featured a dedicated Spanish-language Web site, dubbed Ahora
(www.hp.com/go/hpahora), which offered a sweepstakes contest, product information, a toll-free hotline
and access to Spanish-speaking representatives. An extensive PR effort was also incorporated. HP’s Spanish
speaking employees participated in Demo Days, during which they would go to a retail store and help
customers with products.
The multipronged effort boosted HP’s sales in targeted stores. In areas with integrated efforts HP witnessed a 75 percent year-over-year growth from 2005 to 2006 from holiday to holiday. Apart from the
sales increase, HP also managed to garner good PR buzz. HP’s products were named the “official camera” for five radio stations in Houston, had several on-air mentions in New York and Los Angeles. HP’s
call centers received praise too. Through Spanish-speaking agents, HP had provided the consumers and
prospects with a level of comfort and instant connections. The tremendous success that HP enjoyed
through this integrative marketing approach has made them to consider continuing this initiative further, as against to just a holiday shopping offer.
SOURCE: Nikki Hopewell, “Hewlett-Packard Co.,” Marketing News, November 1, 2007.
7
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
this group is fairly new to the Internet and usage is growing quickly. In June 2003, comScore
Media Metrix reported there were 12.5 million Hispanics online, up 25 percent from 2000. AOL
says it already has 2.3 million Hispanic members, and it wants to be the first-choice ISP for new
online Hispanics, as it has been for so many in the general population. AOL has customized the
e-mail service by providing a greeting in Spanish, which reads “Bienvenidos, tienes e-mail”
instead of “Welcome, you’ve got mail,” that greets the English users. They have also managed
to rope in Coca-Cola, Bank of America, Heineken, L’Oreal, General Mills’ Cheerios, and Toyota
as charter sponsors for AOL Latino.
A.C. Nielsen, on the other hand, has used its Homescan system to analyze Hispanic consumer
habits. Homescan recruits residents of some 500 Latino households who will keep track of their
store purchases and scan the bar codes into la maquinita, or the little machine. Early findings show
that Latino households outspend general market households in many product categories. For
example, the average Hispanic household in the survey spent 67 percent more on carbonated soft
drinks than did the average non-Hispanic household, 89 percent more on fruit drinks, 39 percent
more on cereal, and 41 percent more on toilet paper.
P&G Luring Women with Their Feminine Toothpaste11
The distinction between men’s and women’s grooming aids is very clear, and the last unisex
product left in the bathroom is the toothpaste. Men’s and women’s teeth being the same, it will
take a lot of marketing efforts to make it a success. In 2002 Procter & Gamble Co. launched a
gender-specific toothpaste called Crest® “Rejuvenating Effects” for women.
P&G’s approach was not to make the toothpaste pink or write “for women” on it, but the
rejuvenating effects came in a shimmery box and a hint of vanilla and cinnamon made it a
“feminine” product. For deciding the flavor, researchers gave women about 50 to 100 toothpaste
flavors ranging from mint to vanilla. Women like cinnamon and vanilla flavor the most and also
like sparkle in their paste. Some women suggested a change in the tube so that it would match
with their bathroom accessories. P&G’s target is to have the household buy a number of
toothpastes—for example, one for mom, one for dad, and one for kids.
P&G is desperate for the success of Crest® because it has lost 2/3 of the $1.2 billion market share to Colgate’s “whitening” toothpaste. By making toothpaste for women, who do 82 percent of the grocery shopping, P&G can make the household buy a 25 percent more expensive
product. While testing with men and women, women give a higher rating to rejuvenating effects,
but men did not reject it completely. Men are a little reluctant in using “feminine” products, but
most men go along with whatever product their wives bring home.
The product name “Rejuvenating” stimulates women who are concerned with aging. The
targeted market for P&G is 30–44-year-old women. P&G thinks its new product will help trigger the urgent “rush to brush” feeling in women, and also P&G could make women think about
toothpaste in the same way they do about skin lotion and shampoo.
Harley-Davidson Exploring New Markets12
8
Harley-Davidson, the major U.S. motorcycle maker, was facing problem—stagnant domestic sales.
The management realized that they had to attract more people and extend their appeal beyond
the established base of middle-aged men in order to remain successful. In 1999, Harley-Davidson
started a rental program which provided them a way to hook customers on riding and thereby
entice them into committing to buying a motorcycle. Although the renters were mostly men, the
program served the function of making the buying process simpler for people who had just
learned to ride through Harley’s Rider’s Edge program—40 percent of those enrolled in the program were female and about 30 percent were under the age of 35.
Though the program may have been subtle and not easily measurable, it has shown much
promise from the start. There has been an increase from 6 participating dealers to 250, including 52 in countries such as France, Italy, Canada, Costa Rica, Australia, and Mexico. The company
ROLE
OF
MARKETING RESEARCH
IN
MANAGERIAL DECISION MAKING
intends to expand this further by adding another 20 to 30 participating outlets every year. Motorcycles rented increased from 401 days in 1999 to a total of 224,134 days worldwide in 2004.
Harley-Davidson says that 32 percent of rental customers it surveyed bought a bike or placed
an order after renting, and another 37 percent were planning to buy one within a year. In addition, nearly half of the renters spend more than $100 on Harley-Davidson accessories, such as
T-shirts and gloves.
Harley Davidson has a history riddled with innovative and timely marketing intervention that has allowed Harley-Davidson to sustain its superiority in the premium heavy-weight
motorcycle segment. One of the primary strengths of Harley Davidson is its polarization. People either love a Harley or hate it—and those who do love a Harley really LOVE it. Can you
think of another brand that people actually tattoo to their body? With the Harley Owners
Group (HOG), the company ensures that customers and prospects continue to be active and
enthusiastic. HOG currently has over 295,000 members worldwide, 900 local chapters and is
the largest company-sponsored motorcycle enthusiast group. By organizing a number of
national, touring, and state rallies, the HOG encourages people to use their motorcycle, based
on Harley’s confidence that if people use the motorcycle, they will continually stay involved.
There are thousands of such examples, because virtually every private- and public-sector
organization encounters the same pressures for more and better information about its markets.
Whether the organization serves customers in competitive market environments or clients in a
public-sector enterprise, it is necessary to understand and satisfy the changing needs of diverse
groups of people.
Marketing research is the function that links the consumer, customer, and public to the marketer through
information—information used to identify and define marketing opportunities and problems; generate, refine,
and evaluate marketing actions; monitor marketing performance; and improve understanding of marketing as
a process. Marketing research specifies the information required to address these issues, designs the method
for collecting information, manages and implements the data collection process, analyzes, and communicates
the findings and their implications.
American Marketing Association
Official Definition of Marketing Research
This definition highlights the role of marketing research as an aid to decision making. An
important feature is the inclusion of the specification and interpretation of needed information. Too
often, marketing research is considered narrowly as the gathering and analyzing of data for someone else to use. Firms can achieve and sustain a competitive advantage through the creative use
of market information. Hence, marketing research is defined as an information input to decisions,
not simply the evaluation of decisions that have been made. Market research alone, however, does
not guarantee success; the intelligent use of market research is the key to business achievement. A
competitive edge is more the result of how information is used than of who does or does not have
the information.13
ROLE OF MARKETING RESEARCH IN MANAGERIAL
DECISION MAKING
Marketing decisions involve issues that range from fundamental shifts in the positioning of a
business or the decision to enter a new market to narrow tactical questions of how best to stock
a grocery shelf. The context for these decisions is the market planning process, which proceeds
sequentially through four stages: situation analysis, strategy development, marketing program
development, and implementation. This is a never-ending process, so the evaluation of past
strategic decisions serves as an input to the situation assessment. Figure 1-2 suggests some
elements of each stage. During each stage, marketing research makes a major contribution to
clarifying and resolving issues and then choosing among decision alternatives. The following
sections describe these steps in more detail and describe the information needs that marketing
research satisfies.
9
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
•
Situation
Analysis
•
•
•
Strategy
Development
•
•
Marketing
Program
Development
Implementation
•
•
•
•
•
•
ON
MARKETING INTELLIGENCE
Understand the environment and
the market
Identify threats and opportunities
Assess the competitive position
Define the business scope and served
market segments
Establish competitive advantages
Set performance objectives
Product and channel decisions
Communication decisions
Pricing
Personal selling decisions
Performance monitoring
Refining strategies and program
FIGURE 1-2
Marketing planning process.
Situation Analysis
Effective marketing strategies are built on an in-depth understanding of the market environment of the business, and the specific characteristics of the market. The depth of these
information needs can be seen from the list in Table 1-3, which shows the requirements of a
major consumer packaged-goods manufacturer.
TABLE 1-3
Scope of Situation Assessment for a Consumer Goods Manufacturer
1. Market environment
a. Technologies? How else will customers satisfy their needs?
b. Economic trends? Disposable income?
c. Social trends? What are the trends in age, marital status, working women, occupations, location, and shifts away from the center city? What values are becoming fashionable?
d. Political and regulatory? New labeling and safety requirements.
2. Market characteristics
a. Market size, potential, and growth rate?
b. Geographic dispersion of customers?
c. Segmentation: How many distinct groups are there? Which are growing?
d. Competition? Who are the direct rivals? How big are they? What is their performance? What is
their strategy, intentions, and likely behavior with respect to product launches, promotions, and
the like?
e. Competitive products? Their nature and number?
f. Channel members? What is the distribution of sales through supermarkets and other outlets?
What are the trends? What are they doing to support their own brands?
10
3. Consumer behavior
a. What do they buy? A product or service? A convenience, shopping, or specialty good? A
satisfaction . . . ?
b. Who buys? Everybody? Women only? Teenagers (that is, demographic, geographic, psychographic classification)?
c. Where do they buy? Will they shop around or not? Outlet types?
d. Why do they buy? Motivations, perceptions of product and needs, influences of peers, prestige,
influence of advertising, media?
e. How do they buy? On impulse, by shopping (that is, the process they go through in purchasing)?
f. When do they buy? Once a week? Every day? Seasonal changes?
g. Anticipated change? Incidence of new products, shifts in consumers’ preferences, needs?
ROLE
OF
MARKETING RESEARCH
IN
MANAGERIAL DECISION MAKING
The macroenvironment includes political and regulatory trends, economic and social trends,
and technological trends. Marketing researchers tend to focus on those trends that affect the
demand for products and services. For example, the most important influences on the demand
for consumer packaged-food products over the last decade were
■ Demographic shifts, including a record number of aging adults who were increasingly affluent and active.
■ Rapid changes in family structure as a result of delayed marriages, working wives, and a high
divorce rate.
■ Shifts in values as consumers became preoccupied with their own economic and emotional
well-being.
These trends resulted in increased concerns about the quality of food, nutritional value, personal fitness, and “naturalness.” Equally influential were shifts in food consumption patterns
toward “grazing” or snacking, and more away-from-home eating.
Understanding the customers—who they are, how they behave, why they behave as they
do, and how they are likely to respond in the future—is at the heart of marketing research.
Increasingly, marketing researchers are being asked to turn their talents to understanding the
behavior and intentions of consumers. Marketing Research in Action 1-2 depicts how eating
MARKETING RESEARCH IN ACTION 1-2
Market on the Go
While eating and drinking on the street may seem a normal behavior in the United States, it could
mean violating two fundamental rules in Italy—looking good and eating well. Italians are often taught
that eating or drinking while walking causes indigestion, with the exception of a gelato or a slice of
focaccia being the only food they could enjoy on the street. Italian children too are taught not to eat or
drink anywhere but at the table. This rule is so strictly adhered to, that many shops and businesses
still close for 2 hours between 12:30 and 2:30 in the afternoon so people can enjoy a leisurely meal. This
is also reflected in licensing laws which do not permit sales of food or drinks at places like newsstands.
Italians have always believed in the curative properties of mineral waters, with some small towns
making public fountains out of spring water. According to International Bottled Water Association, a
U.S.-based trade association representing the bottled water industry, per capita consumption of bottled
water in Italy was 53.3 gallons in 2007—third in the world after United Arab Emirates and Mexico, as
compared to the 29.3 gallons in U.S. during the same year. Italy is among the leading producers,
exporters and consumers of mineral water in the world with over 250 brands of Italian mineral water
on the international market and over 600 brands sold domestically. However, Italians drink just 12 liters
a year on the move, half of what Americans do.
Nestle SA, a major player in the bottled-water business, is now trying to build a new market by
changing Italian habits. They have designed a new bottle top with a membrane to prevent leaking
into handbags. To encourage the consumption of bottled water outside, Nestle distributes their water
at soccer games and aerobic sessions.
To promote the “on the go” behavior, Nestle is trying to persuade Italians to drink more bottled water
away from the table through its new offerings. To cater to the young women segment, they have
launched a brand named Acqua Panna. The TV commercials for this squeezable hourglass-shaped 75 cl
(25.4 oz.) bottle, known as Panna 75, featured a fashionable cartoon character named Lulu who carries
a purse containing Acqua Panna, tipped over to show that it does not leak.
Nestle researchers also found that Italian mothers believe that fizzy drinks and ice are not good for
kids. For this, they came up with a brand of still water called Issima—denoting an Italian suffix used in
superlatives—in 11.2 oz. bottles. The marketers at Nestle arranged for free samples to be handed out to
mothers and explained that the squeezable bottle fits neatly in lunch boxes, and is neither fizzy, nor cold,
nor a soda. Nestle’s promotion activities for Issima included conducting contests for kids on its Issima
website with prizes such as radios and flashlights, launching a website that lets kids play a game following the adventures of Issimo—a skateboarding cartoon character—and signing a deal with Autogrill,
Italy’s largest roadside restaurant chain to include Issima in its kids’ meals.
SOURCE: Updated and adapted with permission from Deborah Ball, “Italian Challenge: Water Everywhere, but Not to Go,” The Wall Street Journal, May 23,
2005, p. A.1.
11
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
and drinking habits among Italians have made Nestle in Italy see a new market to promote
their bottled water.
A major responsibility of the marketing research function is providing information that will
help detect problems and opportunities, and then, if necessary, learning enough to make decisions
as to what marketing program would result in the greatest response. An opportunity might be
presented by the sense that customers are increasingly dissatisfied with existing products. Marketing research could be asked to detect the dissatisfaction, perhaps determine how many people
are dissatisfied, and learn the level and nature of that dissatisfaction.
Various research approaches are used to analyze the market. Perhaps the simplest is to organize information already obtained from prior studies, from magazine articles that have been
printed, and from customers’ comments to a firm’s sales representatives. Another approach is to
have small groups of customers, called focus groups, discuss their use of a product. Such discussion groups can provide many ideas for new marketing programs.
When a problem or opportunity has been identified and it is necessary to understand it in
greater depth, a survey is often employed. For example, to understand the competitive position
of Quebec in the tourism market, a survey was conducted to determine the benefits sought by
visitors and nonvisitors, as well as the risks they perceived. The results identified a large group
who felt highly insecure in new and/or foreign environments and were not attracted by the
appeals of uniqueness in culture, traditions, and architecture that Quebec used to differentiate
its product. Marketing Research in Action 1-3 is a prime example of a company repackaging and
tailoring its product range based on customer preferences and demand.
Strategy Development
12
All businesses have strategies, which are the methods used to make and sell products or perform services. Often, strategies are determined by a company’s reaction to events beyond its control rather than by solid marketing intelligence and strategic planning. But, the question asked
is why do firms plan? The answer is simple: competitive advantage.14
Several factors contribute to organizational growth in both size and complexity. Decision
making gets more and more complex as the size of the business and market share increases, and
vice versa when the business size and market share decreases. This implies a critical need for
strategic focus—focusing on customer/competitive analysis. Both elements are critically dependent on rigorous marketing intelligence.15
In order for companies to maximize opportunity, they must first assess their strategic position.
Only then will management be able to decide where and how the company should position itself.
Evaluations of past performance, marketing strengths and weaknesses, reputation for quality products, utilization issues, and mission need to be addressed. All of these issues can be addressed by
strategic planning and good marketing intelligence.16
During the strategy development stage the management team of the business decides on
answers to three critical questions. Marketing research provides significant help in finding the
answers to these questions.
1. What business should we be in? Specifically, what products or services should we offer?
What technologies will we utilize? Which market segments should we emphasize? What channels should we use to reach the market? These are far-reaching choices that set the context for
all subsequent decisions.
These questions have become especially compelling in markets that are mature and saturated, including not only most packaged goods but also household appliances, automobiles, and
services such as banking and air travel. One sure route to growth in this competitive environment is to create highly targeted products that appeal to the tastes of small market segments.
Research supports this search for niches with large-scale quantitative market studies that describe
buying behavior, consumer beliefs and attitudes, and exposure to communications media. Large
samples are needed to delineate the segments, indicate their size, and determine what the people in each segment are seeking in a product.
ROLE
OF
MARKETING RESEARCH
IN
MANAGERIAL DECISION MAKING
MARKETING RESEARCH IN ACTION 1-3
Restructuring and Rejuvenating McDonald’s
With more than 35,000 stores in 100 countries, McDonald’s is the world’s largest fast-food chain. Despite
a whopping 43 percent market share, McDonald’s sales had been stagnant for several years and dipped
2.8 percent in the third quarter of 2002. This spurred a dramatic turnaround in corporate focus, leading to
the adoption of a two pronged strategy. In the United States, McDonald’s has focused on increasing sales volume and quality at existing locations by renovating stores, increasing store hours and expanding menu
offerings. Internationally, McDonald’s has expanded aggressively through franchisees.
Despite its size, these strategies have sufficiently paid off in terms of sales and revenue. As of 2008,
McDonald’s has been responsible for 19% of the total revenue earned in the Fast-Food industry, and possess approximately 90% market share in the 67 billion dollar Fast-Food Hamburger Restaurant sector, In
the local market, McDonald’s does not seem to show any signs of slowing down even in tough economic
situations and falling consumer spending, and is even benefiting as consumers trade down from more
expensive eating options, while international operations continue to drive profit growth. Explosive growth
of a global middle class, particularly in emerging markets like China, India and Latin America, has been
a massive opportunity for McDonald’s.
McDonald’s has effectively used its size to muscle out competition in three primary ways:
■ Uniform menu offerings can be mass produced, lowering production costs.
■ Bargaining power with suppliers lowers input costs and boosts margins.
■ Large advertising budget means lots of domestic and international exposure.
SOURCE: Adapted and updated from “McDonald’s Trying to Get Out of The Same Old Grind,” Wall Street Journal, November 3, 2002.
2. How will we compete? Next the management team has to decide why the business is better than the competition in serving the needs of the target segment, and what has to be done to
keep it in front. Competitive superiority is revealed in the market as either differentiation along
attributes that are important to target customers, or the lowest delivered cost position. Otis Elevator is able to dominate the elevator business by using information technologies to provide
superior service response and preventive maintenance programs that reduce elevator breakdowns—
attributes that customers appreciate.
Marketing research is essential for getting answers to three key questions about differentiation: What are the attributes of the product or service that create value for the customer? Which
attributes are most important? How do we compare to the competition? For example, every
movie has a dozen different story lines, and Joseph Helfgot, a sociologist turned Hollywood market researcher, says he knows which one you want to hear. Helfgot is the one who tells studio
executives how to sell their movies and sink-or-swim release schedules. Consider The Silence of
the Lambs (Helfgot won’t discuss any of his recent projects). Is it about a gruesome serial murderer? Or a feisty FBI ingenue? Or a brilliant psychotic helping the FBI catch a like-minded fiend?
None of the above, Helfgot found out. Audiences were interested in hearing a story about the
bizarre relationship between Jodie Foster, an FBI academy graduate, and imprisoned serial killer
Hannibal Lecter—played by Anthony Hopkins—who helped her solve the murders. That became
the focus of the publicity campaign for what turned out to be a blockbuster hit. With marketing
costs accounting for up to a third of movie budgets and with competition for space in movie
theaters becoming increasingly fierce, studios have come to rely on whatever information science
can provide to help sell their movies.
The attributes of value go well beyond physical characteristics, to encompass the support
activities and systems for delivery and service that make up the augmented product. In the
lodging market, the key attributes are honoring reservations on time, providing good value for
the money, and the quality and amenities of the guest rooms. Each market has unique attributes that customers employ to judge the competitive offerings, which can be understood only
through careful analysis of usage patterns and decision processes within that market. This
knowledge comes from informed sources and in-depth customer surveys.
13
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
An understanding of competitive advantage also requires detailed knowledge of the capabilities, strategies, and intentions of present and prospective competitors. Marketing research contributes here in two ways: identifying the competitive set, and collecting detailed information
about each competitor. Some ways of undertaking competitive intelligence work are discussed
later in the book. Marketing Research in Business 1-4 talks about how marketing research helped
a speech-recognition software provider find his market.
3. What are the objectives for the business? An objective is a desired performance result that
can be quantified and monitored. There are usually objectives for revenue growth, market share,
and profitability. Increasingly, firms are adopting objectives for service levels (for example, speed
of response to quotations) and customer satisfaction. Marketing research is needed to establish
both the market share and the level of customer satisfaction. Sometimes share information—we
have x percent of the y market—is readily available from secondary sources, but it may not be
if the served market is different from the standard definition, or if share is defined in dollar sales
terms rather than unit volume.
Marketing Program Development
Programs embrace specific tasks, such as developing a new product or launching a new advertising campaign (see Marketing Research in Action 1-4). An action program usually focuses on a
single objective in support of one element of the overall business strategy. This is where the bulk
of ongoing marketing research is directed. An idea of the possibilities of and needs for research
can be gathered from Table 1-4, which describes some of the representative program decisions
that utilize information about market characteristics and customer behavior.
To illustrate some of the possible research approaches that are employed, we will focus on
the series of market research studies that were conducted to help Johnson Wax Company successfully introduce Agree Creme Rinse in 1977 and Agree Shampoo in 1978. The story begins
MARKETING RESEARCH IN ACTION 1-4
Understanding a Specialized Market
In 1998, a group of software students at Carnegie Mellon University, Pittsburgh developed speechrecognition software. As the software took account of stammers, pauses, and other quirks for increasing
the accuracy of its transcriptions, they believed they had a great business idea. They launched a company called Multimodal Technologies and signed on some customers by licensing the software for uses
such as video game interfaces, customer call centers, and automakers. However, the voice-recognition
software did not fit the customers well. In the tests conducted, customers were not able to produce documents with the accuracy that the programmers could manage. They realized that they had built an
application which no one could work with.
The business model was discarded by Multimodal and it started concentrating on a narrower market. It
started by changing focus from a software seller through licensing to a service provider and began customizing the software for the customers to better suit their needs. It narrowed in on the health-care market
as its prime focus, recognizing the industry’s need for cost-saving technologies. It then began pitching for
the University of Pittsburgh Medical Center (UPMC), which spends more than $20 million a year to transcribe the diagnoses and prescriptions that doctors dictate into tape recorders. After months of negotiation,
a pilot test was launched in 2003 in the hospital’s radiology department. The results showed accuracy to be
95 percent and savings totaled about 40 percent.
Apart from addressing the transcription issues, the software also finds its uses in converting digital
documents into different formats for patients, insurers, and administrators; easier search functions by
locating more key words among others. In all, the software provides more advantages like greater savings and efficiencies than just reducing the labor of transcribing dictation. In the words of Michael Finke,
founder and chief executive of Multimodal Technologies, “Nobody buys you because you are a cool technology; you had better solve someone’s problem.”
14
SOURCE: Adapted with permission from Paulette Thomas, ”Case Study: For Profits, Technology Must Answer a Need,” The Wall Street Journal—Eastern
Edition, October 26, 2004, p. B. 4.
ROLE
OF
MARKETING RESEARCH
IN
MANAGERIAL DECISION MAKING
TABLE 1-4
Developing the Marketing Program—Representative Decisions
That Draw on Marketing Research
1. Segmentation decisions
Which segment should be the target?
What benefits are most important for each segment?
Which geographic area should be entered?
2. Product decisions
What product features should be included?
How should the product be positioned?
What type of package is preferred by the customers?
3. Distribution decisions
What type of retailer should be used?
What should be the markup policy?
Should a few outlets be employed or many?
4. Advertising and promotion decisions
What appeals should be used in the advertising?
In which vehicles should the advertising be placed?
What should the budget be?
What sales promotion should be used, and when should it be scheduled?
5. Personal selling decisions
What customer types have the most potential?
How many salespeople are needed?
6. Price
What
What
What
decisions
price level should be changed?
sales should be offered during the year?
response should be made to a competitor’s price?
7. Branding decisions
What should be the name, symbol, logo, and slogan that will be associated with the product?
What is the position that the brand should adopt vis-à-vis the competition?
How can brand loyalty be increased?
8. Customer satisfaction decisions
How should customer satisfaction be measured?
How often should it be measured?
How should customer complaints be handled?
with a major market analysis survey of hair-care practices that was conducted in the early 1970s.
The study showed that there was a trend away from hair sprays, but a trend toward shampooing hair more frequently and a growing concern about oily hair. This led to a strategic decision
to enter the shampoo and creme rinse market with products targeted toward the oiliness problem. This decision was supported by other studies on competitive activities in the market and
on the willingness of the retailers to stock new shampoos.17
A total of 50 marketing research studies conducted between 1975 and 1979 supported the
development of these two products. A series of focus group discussions was held to understand
the oiliness problem and people’s perceptions of existing shampoo products. The firm was particularly interested in learning about teenagers, since most of its products were sold to homemakers. One goal of these focus groups was to get ideas for a copy theme. Subsequently, more
focus groups were held to get reactions to the selected advertising theme, “Helps Stop the
Greasies.” Several tests of advertising were employed in which customers were exposed to advertisements, and their reactions were obtained. In fact, more than 17 television commercials were
created and tested.
15
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
More than 20 of the studies helped to test and refine the product. Several blind comparison
tests were conducted, in which 400 women were asked to use the new product for two weeks
and compare it to an existing product. (In a blind test the products are packaged in unlabeled
containers and the customers do not know which contains the new product.)
Several tests of the final marketing program were conducted. One was in a simulated supermarket where customers were asked to shop after they had been exposed to the advertising. The
new product, of course, was on the shelf. Another test involved placing the product in an actual
supermarket and exposing customers to the advertising. Finally, the product was introduced
using the complete marketing plan in a limited test area involving a few selected communities
including Fresno, California, and South Bend, Indiana. During the process, the product, the
advertising, and the rest of the marketing program were being revised continually. The effort
paid handsome dividends: Agree Creme Rinse took a 20 percent share of the market for its category and was number one in unit volume, and Agree Shampoo also was introduced successfully.
In the 1980s, Gillette introduced Mink Difference Shampoo, containing mink oil, for the older
market, and for the younger segment, Silkience, a self-adjusting shampoo that provided differential conditioning depending on the user’s hair type.
Implementation
The beginning of the implementation phase is signaled by a decision to proceed with a new program or strategy and by the related commitments to objectives, budgets, and timetables. At this
point the focus of marketing research shifts to such questions as
Did the elements of the marketing program achieve their objectives?
How did sales compare with objectives?
In what areas were sales disappointing? Why?
Were the advertising objectives met?
Did the product achieve its distribution objectives?
Are any supermarkets discontinuing the product?
Should the marketing program be continued, revised, or expanded?
Are customers satisfied with the product?
Should the product be changed? More features added?
Should the advertising budget be changed?
Is the price appropriate?
For research to be effective at this stage, it is important that specific measurable objectives
be set for all elements of the marketing program. Thus, there should be sales goals by geographic
area; distribution goals, perhaps in terms of the number of stores carrying the product; and
advertising goals, such as achieving certain levels of awareness. The role of marketing research
is to provide measures against these objectives and to provide more focused studies to determine why results are below or above expectations.
Often underlying this phase of marketing management is uncertainty about the critical judgments and assumptions that preceded the decision. For example, in the first half of 1996, Compaq,
IBM, Dell, and Gateway all launched new products in the personal computer market. Compaq
introduced its Presario line of personal computers, designed to target specific groups of consumers.18 Prior to this, most companies targeted only one brand of PC toward the home segment.
One reason was a fundamental assumption about whether customers use computers for purposes other than traditional computing tasks. In response to this uncertainty, the companies
undertook research studies to measure the acceptability of new product entries and of maintaining computer usage at home.
There is overlap among the phases of the marketing process. In particular, the last phase, by
identifying problems with the marketing program—and perhaps opportunities as well—eventually
blends into the situation analysis phase of some other follow-up marketing program.
16
FACTORS THAT INFLUENCE MARKETING RESEARCH DECISIONS
FACTORS THAT INFLUENCE MARKETING
RESEARCH DECISIONS
Marketing research is not an immediate or an obvious path to finding solutions to all managerial problems. A manager who is faced with a particular problem should not instinctively resort
to conducting marketing research to find a solution to the problem. A manager should consider
several factors before ordering marketing research and conduct a value analysis to answer the
simple question—will the reduction in risk be greater than the cost?—which includes the time
delay associated with waiting to make the decision. In many situations it is best not to conduct
marketing research. Hence, the primary decision to be made is whether or not market research
is called for in a particular situation. The factors that influence this initial decision include the
following.
Relevance
Research should not be conducted to satisfy curiosity or confirm the wisdom of previous decisions. Relevance comes through support of strategic and tactical planning activities, that is, by
anticipating the kinds of information that will be required. This information is the backbone of
the ongoing information system. As new circumstances arise and decision alternatives become
more specific, research projects may be undertaken. Throughout the planning of these projects,
the focus must be constantly on decisions.
Type and Nature of Information Sought
The decision whether to conduct marketing research depends on the type and nature of the
information sought. If the information required for decision making already exists within the
organization, in the form of results of a study conducted for a different problem or in the form
of managerial experience and talents, marketing research is not called for. Under these circumstances, further research would be redundant and a waste of money. For example, Procter &
Gamble, using its prior knowledge of the U.S. coffee market, launched Folger’s Instant Coffee
nationally, after some preliminary research. The same is true for many organizations that have
accumulated rich experience in a particular market and already possess the information
required to solve a certain problem.
Timing
Research decisions are constrained by the march of events. Often these decisions are fixed in time
and must be taken according to a specified schedule, using whatever information is available.
If a new product is to be launched in the spring, all the research-based decisions on price,
product formulation, name, copy appeals, and other components must be conducted far in
advance. One role of the planning system is to schedule needed market research so that it can
be conducted in time to influence decisions. The formulation of responses to competitive
actions puts the greatest time pressure on researchers, for the results are always wanted “yesterday.” There are, of course, many situations where the timing of decisions is contingent upon
the research results. Even so, there is still time pressure stemming from the recognition that
failure to take corrective action or pursue an opportunity as quickly as possible will result in
opportunity costs.
Availability of Resources
Though the need for resources to be available may appear to be obvious, in several instances
managers have called for marketing research without properly understanding the amount of
17
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
resources available—including both financial and human resources. Lack of funds can result in
improper and inefficient execution of a marketing research project. The results of such research
often will be inaccurate. Again, if funds are available to conduct proper research but are insufficient to implement the results of the research, the marketing research is made useless. Also, the
availability of a talent pool is a critical issue in deciding whether or not to conduct extensive
marketing research. This is particularly so when the research is being conducted by an external
source. When poorly qualified researchers are hired, the weaknesses in their training and lack of
insight produce unimpressive and often inapplicable results.
Cost-Benefit Analysis
Before conducting marketing research, a prudent manager should perform a cost–benefit analysis to determine the value of the information sought through the research. Willingness to acquire
additional decision-making information by conducting marketing research depends on a manager’s perception of the incremental quality and value of the information vis-à-vis its cost and
the time it would take to conduct the research. Hence, before conducting marketing research, it
is necessary to have some estimate of the value of the information being sought. Such an estimate will help determine how much, if anything, should be spent on the research.
USE OF MARKETING RESEARCH
Although research is conducted to generate information, managers do not readily use the information to solve their problems. The factors that influence a manager’s decision to use research
information are (1) research quality, (2) conformity to prior expectations, (3) clarity of presentation, (4) political acceptability within the firm, and (5) challenge to the status quo.19 Researchers
and managers agree that the technical quality of research is the primary determinant of research
use. Also, managers are less inclined to utilize research that does not conform to prior notions
or is not politically acceptable.20 Some researchers argue that the use of information is a function of the direct and indirect effects of environmental, organizational, informational, and individual factors.21 However, a researcher should not alter the findings to match a manager’s prior
notions. Further, managers in consumer organizations are less likely to use research findings than
their counterparts in industrial firms.22 This is due to a greater exploratory objective in information collection, a greater degree of formalization of organizational structure, and a lesser
degree of surprise in the information collection.
Does Marketing Research Guarantee Success?
It is easier to conduct research and generate information than to understand the consequences
of the information. Many companies with excellent marketing research experience have failed
in their efforts to capture the actual needs of the consumers. For example, Coca-Cola conducted numerous studies before introducing New Coke. The study results revealed that New
Coke tasted better than the original Coke, yet the product failed in the marketplace because
of consumers’ strong emotional/loyalty attachment to the original Coke.23 Realizing that the
market needed a low-calorie beer, Gablinger introduced the first low-calorie beers. However,
the poor taste of the beer led to its downfall. Later, Anheuser-Busch and Miller Brewing
achieved great success by emphasizing the benefits of good taste and a less filling product
(rather than fewer calories, although what the market wanted was a low-calorie beer).24 Findings in international markets have shown that a product should be priced in a manner in
which the customer will realize the value received, and that the price should be within reach
18
ETHICS
IN
MARKETING RESEARCH
of the target market. This was observed in the case of Warner-Lambert’s launch of its fiveunit pack of Bubbaloo bubble gum in Brazil. Despite bubble gum representing a majority of
the overall gum sector in Brazil, the product failed. The reason was the high price with respect
to the target market. The company then re-launched the gum as a single-unit pack and revised
the price within the range of the target segment. This price revision stabilized the brand and
helped improve sales.25
ETHICS IN MARKETING RESEARCH
Ethics refers to moral principles or values that generally govern the conduct of an individual or
group. Researchers have responsibilities to their profession, clients, and respondents, and must
adhere to high ethical standards to ensure that both the function and the information are not
brought into disrepute. The Marketing Research Association, Inc. (Chicago, Illinois) has instituted
a code of ethics that serves as a guideline for marketing ethical decisions. Marketing Research
in Action 1-5 shows the Code of Professional Ethics and Practices instituted by the Marketing
Research Association.26 The Council of American Survey Research Organization (CASRO) has
also established a detailed code of marketing research ethics to which its members adhere.27 Normally, three parties are involved in a marketing research project: (1) the client who sponsors the
project, (2) the supplier who designs and executes the research, and (3) the respondent who provides the information. The issue of ethics in marketing research involves all three players in a
research project. In a recent case, both the sponsor’s and the supplier’s ethics were found to be
in question. Marketing Research in Action 1-6 provides the details of the case.
MARKETING RESEARCH IN ACTION 1-5
Code of Ethics of Marketing Research Association
The Code of Professional Ethics and Practices
1. To maintain high standards of competence and integrity in marketing and survey research.
2. To maintain the highest level of business and professional conduct and to comply with Federal, State and
local laws, regulations and ordinances applicable to my business practice and those of my company.
3. To exercise all reasonable care and to observe the best standards of objectivity and accuracy in the development, collection, processing and reporting of marketing and survey research information.
4. To protect the anonymity of respondents and hold all information concerning an individual respondent
privileged, such that this information is used only within the context of the particular study.
5. To thoroughly instruct and supervise all persons for whose work I am responsible in accordance with
study specifications and general research techniques.
6. To observe the rights of ownership of all materials received from and/or developed for clients, and to
keep in confidence all research techniques, data and other information considered confidential by their
owners.
7. To make available to clients such details on the research methods and techniques of an assignment as
may be reasonably required for proper interpretation of the data, providing this reporting does not violate the confidence of respondents or clients.
8. To promote the trust of the public for marketing and survey research activities and to avoid any procedure which misrepresents the activities of a respondent, the rewards of cooperation or the uses of data.
9. To refrain from referring to membership in this organization as proof of competence, since the organization does not so certify any person or organization.
10. To encourage the observance of principles of this code among all people engaged in marketing and survey research.
Visit http://www.marketingpower.com/Pages/default.aspx for more information on this organization.
SOURCE: Reprinted by permission of the Marketing Research Association, Inc., Chicago, IL.
19
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
MARKETING RESEARCH IN ACTION 1-6
Unilever falls victim to ‘dumpster diving’
Procter & Gamble and Unilever, the leading consumer packaged goods companies are known for their
fierce competition. In 2001, Procter & Gamble hired private investigators to find more about Unilever’s
hair care business.
The investigators had sifted through rubbish bins outside Unilever’s offices and succeeded in gathering piles of unshredded documents relating to Unilever’s plans for the shampoo market. While there
are no clear rules on the legal status of rubbish, the practice commonly known as ‘dumpster diving’
broke Procter & Gamble’s own internal guidelines on intelligence gathering. When Procter & Gamble
were alerted regarding this practice, the management claimed that a few ‘rogue operators’ had apparently overstepped the mark in their eagerness to provide high level intelligence.
Following this breach, Unilever responded by threatening legal action against their rivals, and the case
was finally resolved when Procter & Gamble made an out of court settlement of around $10 million.
SOURCE: Andrew Crane, “In the company of spies: the ethics of industrial espionage,” July 2003, International Center for Corporate Social Responsibility,
ISSN 1479-5124.
THE RESPONDENT’S ETHICS AND RIGHTS
A respondent who of his or her own free will agrees to participate in a marketing research project has the ethical obligation to provide the supplier, and hence the client, with honest and
truthful answers. The respondent can abstain from answering a sensitive question, but falsifying the answer is ethically improper.
Any respondent who participates in a research project has the following rights:
■ The right to privacy.
■ The right to safety.
■ The right to know the true purpose of the research.
■ The right to the research results.
■ The right to decide which questions to answer.
The Sponsor’s Ethics
The sponsor, or the research client, has to abide by a number of ethical or moral rules and regulations when conducting a research study. The more common sources of ethical problems in the client
establishment stem from the following sources.
Overt and Covert Purposes. Most researchers have encountered situations where the main purpose
of their efforts was to serve someone’s organizational goals. Thus, research can be used to postpone
an awkward decision or to lend respectability to a decision that has been made already. A related
purpose is to avoid responsibility. When there are competing factions, the manager who must make
a difficult choice looks to research to guide the decision. This has the further advantage that if the
decision is later proven wrong, the manager can find someone else to blame.
Sometimes a covert purpose will open the way to ethical abuses that present difficult dilemmas to researchers. The most serious abuses are created when there is subtle (or not so subtle)
pressure to design research to support a particular decision or enhance a legal position.
20
Dishonesty in Dealing with Suppliers. A few client companies have been known to indulge in
“picking the brains” of research suppliers by asking them to submit elaborate bids that detail
the research design and methodology the supplier would adopt in conducting the research. Later,
the client firm uses these ideas and conducts the research on its own. Another technique that
client firms sometimes use is to make a false promise of future contracts in an effort to obtain a
low price for the current project.
THE RESPONDENT’S ETHICS
AND
RIGHTS
Misuse of Research Information. The client firm should not misuse information gathered through
marketing research projects. For example, databases about consumer preferences are used in target marketing to identify the people who are most likely to buy or use a product. Blockbuster
Entertainment Corporation (BEC), the world’s largest video rental chain, has about 9,100 companyowned or franchised stores in 25 countries, with around 6,000 in the United States.28 The company
rents more than 1 billion videos, DVDs, and video games at its Blockbuster Video outlets each year.
Using sophisticated computer-matching information technologies, BEC has also formed a massive
database of information on the movie videos it rents, and the information can be made available
for resale to direct marketers.29 Even though a federal law prohibits release of the names of video
renters, there is no rule against divulging the subject matter of those movies. BEC even made public its intent to sell its renter data,30 which sparked a debate over consumer privacy. BEC soon came
under Federal Trade Commission vigilance, faced a torrent of customer criticism, and received considerable negative publicity. Even though BEC does not rent X-rated movies, the concern and controversy that BEC had infringed on customer privacy rights reached a feverish pitch. The common
form of misuse comes from comparative advertisements or product performance claims that stem
from results of tests done on data that are statistically not significant. Though puffery in advertisements is a normal practice, gross misuse of research data is ethically unacceptable.31 Marketing Research in Action 1-7 provides the current legislative scenario regarding disposal of data.
Too often researchers find themselves dealing with demands by sales or other personnel for
access to results and the names and telephone numbers of respondents. The intention, of course,
is to use the research study for the entirely different—and usually unethical—purpose of generating sales leads. The only time this is acceptable is when the interviewer asks specifically
whether the respondent will accept a follow-up sales call, or would like more information, and
acts precisely on the respondent’s answer.
MARKETING RESEARCH IN ACTION 1-7
Data Disposal
CMOR (Council for Marketing and Opinion Research) is an organization that deals with state and federal legislation that affects market and opinion research. Some of its tasks include monitoring telephone
solicitation bills, electronic monitoring legislation, data privacy, consumer list legislation, caller ID/call
blocking, and political telemarketing.
According to CMOR, personal or client information discarded in the trash, any data stored on discarded or donated computer technology (such as hard drives and thumb drives), any data stored on
devices that are thrown away or donated to charity, is legally and effectively open to anyone. Further,
electronic data kept beyond its usefulness invites mischief or accidental breach. Some of the reasonable
measures could include establishing and complying with policies to:
■ Burn, pulverize, or shred papers and destroy or erase electronic files or media containing personal
data so that the information cannot be read or reconstructed. This includes any items destined for
recycling, charity, or some other form of use;
■ Only maintain data as long as necessary;
■ Audit your organization’s data holdings on a regular basis and keep track of what data is disposed
when—and how;
■ Conduct due diligence and hire a document destruction contractor to dispose of material specifically
identified as personal data.
Some of the due diligences that can be adopted by researchers are:
■ Reviewing an independent audit of a disposal company’s operations and/or its legal compliance;
■ Obtaining information about the disposal company from several references;
■ Requiring that the disposal company be certified by a recognized trade association;
■ Reviewing and evaluating the disposal company’s information security policies or procedures.
For more information on CMOR, go to www.cmor.org
SOURCE: http://www.mra-net.org/ga/documents/bestpractices/datadisposal_mra_best_practices.pdf
21
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
Disguising Selling as Research. Sadly, there are a number of situations where the research study
is simply a disguise for a selling pitch. Many people have received phone calls, ostensibly to ask
some research questions, that lead only to a canned sales message for life insurance, an encyclopedia, or a mutual fund. This is not only unethical behavior because it has no merits on its own,
it is also a serious abuse of respondent rights. Not surprisingly, respondents are more suspicious
after a few of these encounters and may refuse to participate in any research study.
The Supplier’s Ethics
The more common ethical issues for the research supplier are
■ Violating client confidentiality. Disclosing any information about the client that the supplier has
gathered from the research project amounts to a violation of client confidentiality.
■ Improper execution of research. Suppliers are required to conduct marketing research projects in
an objective manner, free from personal biases and motives. Improper execution also includes
using biased sampling, ignoring relevant data, or misusing statistics, all of which lead to erroneous and misleading results. Marketing Research in Action 1-5 gives a brief listing of federal
and state legislation that affects market and opinion research.
Abuse of Respondents
Abuse of respondents is perhaps the most frequent and controversial problem that crops up
regarding ethics in conducting research. Any form of violation of a respondent’s rights amounts
to unethical treatment or abuse of the respondent.
INTERNATIONAL MARKETING RESEARCH
22
The increase in international trade and the emergence of global corporations resulting from
increased globalization of business have had a major impact on all facets of business, including
marketing research. The increase in global competition, coupled with the formation of regional
trading blocs such as the European Community (EC) and the North American Free Trade Agreement (NAFTA), have spurred the growth of global corporations and the need for international
marketing research. The need to collect information relating to international markets, and to monitor trends in these markets, as well as to conduct research to determine the appropriate strategies
that will be most effective in international markets, is expanding rapidly. For example, HEB, the
largest grocery store chain in Texas, observed that consumers from Mexico crossed the border to
shop at HEB stores in Texas. After conducting some marketing research in Mexico, HEB decided
to open its first store (and the first American supermarket) in Monterrey, Mexico. The stores will
be somewhat different from their Texas counterparts, with 70 to 80 percent of the products purchased in Mexico and the remainder in the United States. Research revealed that awareness of
HEB is quite high among Mexican consumers and that they place a great deal of emphasis on American products (which are better from the consumers’ perspective).32 HE Butt Grocery is a regional
supermarket chain, operating 300 stores throughout the Southwest, primarily in Texas. The company’s H-E-B Plus stores range in size from 110,000 to nearly 200,000 square feet. HE Butt also operates specialty stores Central Market (gourmet foods) and Mi Tienda (Hispanic foods), and has
29 stores in Mexico. It employs around 70,000 employees and grossed a revenue of over $15 billion
in 2007.33 Wal-Mart acquired Cifra, Mexico’s largest supermarket chain, in 1997 and changed its name
to Wal-Mart de Mexico in 2000.34 So many signs are positive for Mexico’s economy.35
The marketing research industry in the United States is increasingly growing into an international industry, with more than one-third of its revenues coming from foreign operations. The
increase in the importance of global business has caused an increase in awareness of the problems related to international research. International marketing research can be defined as marketing research conducted either simultaneously or sequentially to facilitate marketing decisions
INTERNATIONAL MARKETING RESEARCH
in more than one country.36 As such, the basic functions of marketing research and the research
process do not differ in domestic and multicountry research; however, the international marketing research process is much more complicated and the international marketing researcher faces
problems that are different from those of a domestic researcher. Marketing Research in Action 1-8
talks about Coca-Cola’s foray and experience in the Chinese market.
Throughout this book, we will be discussing the international aspect of the marketing
research process, and, when applicable, we will be highlighting the differences between domestic and international research.
The task of marketing research is to find a sizable segment with homogeneous tastes. The
growing presence of an international market in the United States has been influenced by both
domestic and foreign markets. In the domestic arena, ethnic groups range from Chinese to Turkish, each lending a piece of its culture to the U.S. market. Within each ethnic group, the product
preference is diverse. These facts present a challenge to marketing research to find a homogeneous
group among the “melting pot” of international products. Complicating matters is the rise in foreign manufacturers selling their goods in the United States. Marketing Research in Action 1-9
addresses the issue of viewing U.S. consumers as global consumers.
MARKETING RESEARCH IN ACTION 1-8
Coca-Cola Charting Success in China
In an effort to offset slower growth at home, Coca-Cola has been aggressively expanding in international
markets. When it entered China in 1979, the carbonated drinks market was small and dominated by local
brands. It created a strategy of partnering with and acquiring local bottling companies played an important role in its capturing the Chinese market. The key to Coca-Cola’s success in China can be attributed
to its “high-road” brand positioning strategy.
Focus on innovation to bring back customers for more forms the crux of this high-road strategy. While
the carbonated beverages category existed in China before Coca-Cola, the market was small and consisted of a few low-end carbonated fruit drinks, which were sold mainly to tourists. Coca-Cola created
the carbonated drinks category as it exists today by introducing products new to the Chinese beverage
market, starting with its flagship Coke.
Coca-Cola’s Fanta and Smart—flavors that were launched exclusively for the Chinese market—became
top earners for the company. The company continues to launch new products and has branched out into
related product categories such as beverages, bottled water and fruit juices.
Coca-Cola has always supported new product launches with aggressive marketing initiatives. The
development of a Chinese version of their logo and the usage of celebrity endorsements are some examples of their customer-focused outlook and their interest in the Chinese market. Their strategy is clearly
producing results—over the past decade, its sales have grown at over 20 percent in a category growing
at only half that rate.
For more information on Coca-Cola products go to http://www.coca-cola.com.
SOURCE: Ann Chen and Vijay Vishwanath, “Be the Top Pick in China,” Business Times Singapore, January 16, 2004.
MARKETING RESEARCH IN ACTION 1-9
Are U.S. Consumers Global Customers?
A typical day in the life of an American: “He drove his German car made of Swedish steel and interior
of Argentine leather to a gasoline station, where he filled up with Arab oil shipped in a Liberian tanker
and bought two French tires, composed of rubber from Sri Lanka. At home, he dropped his Moroccan
briefcase, hung up his Scottish tweed coat, removed his Italian shoes and Egyptian cotton shirt, then
donned a Hong Kong robe and matching slippers from Taiwan. More comfortable now, he poured a cup
of hot Brazilian coffee into an English coffee mug, set a Mexican place-mat on an Irish linen tablecloth
atop a Danish table varnished with linseed oil from India. Then he filled his Austrian pipe with Turkish tobacco, lit it, and picked up a Japanese ballpoint pen with which he wrote a letter to his congressman demanding to know why the United States has an unfavorable balance of trade.”
SOURCE: http://www.military-quotes.com/forum/only-america-t65609.html
*Original source unavailable.
23
CHAPTER 1 A DECISION-MAKING PERSPECTIVE
ON
MARKETING INTELLIGENCE
Summary
Marketing research links the organization with its market environment. It involves the specification, gathering, analysis, and interpretation of information to help management understand that particular market
environment, identify its problems and opportunities, and develop and evaluate courses of marketing
action. The marketing management process involves situation analysis, strategy development, and marketing program development and implementation. Each of these areas includes a host of decisions that
need to be supported by marketing research information. Marketing research, to be effective, should be
relevant, timely, efficient, accurate, and ethical.
Questions and Problems
1.
2.
3.
4.
5.
6.
How might the following use marketing research?
Be specific.
a. A small sporting goods store
b. Continental Airlines
c. Ohio State University
d. Houston Astros baseball team
e. Sears Roebuck
f. A major television network (CBS, NBC, or ABC)
g. Compaq Computers
h. A museum in a major city
i. A shopping mall in downtown Houston
How might marketing research be used to support
each of the steps in Figure 1-1 that describes the
marketing planning process? For example, how
could it help select the served market segment?
What are some ethical problems that marketing
researchers face in designing and conducting field
studies?
In some companies, strategic planning and marketing research functions both report to the same
executive and may be more or less integrated.
What are the advantages of locating the research
function in this part of the organization? What
arguments could be made in opposition to this
arrangement?
Most companies have entire marketing research
studies, or portions of entire studies, such as interviewing, done by outside suppliers. What factors
will determine whether a firm decides to “make
versus buy”—that is, to contract out most or all of
a study or conduct it themselves?
How does marketing research directed toward
strategy development differ from that directed
toward marketing program development?
7.
8.
9.
Fred Burton, the owner of a small tennis club
in Wichita, Kansas, feels that a demand exists
for indoor courts that presently is not being served.
He is considering employing a marketing research
company to conduct a study to ascertain whether a
market exists for the indoor facilities.
a. What factors should Mr. Burton consider before
ordering market research to be conducted?
b. What are the possible pitfalls that the marketing
research company must avoid while conducting
the study?
c. After obtaining the market research recommendations, Mr. Burton decides not to use the information generated by the market research study.
What factors could have influenced his decision
not to use the research information?
Linda Phillips, an engineering student, has
designed an innovative piece of equipment to
help the physically disabled to communicate. The
equipment incorporates a system of electronic
signals emitted with a slight turn of the head. She
feels that this product could have commercial success if marketed to health-care organizations, but
she has had no past experience in marketing management and does not know how to undertake
the market planning and evaluation process.
Acting as Ms. Phillips’s marketing consultant,
suggest a course of action to help her bring this
innovative product to its market.
Despite the presence of a written code of ethics,
why are some marketing ethical problems hard to
cure? Discuss this from both the sponsor and supplier points of view.
End Notes
1. http://www-01.ibm.com/software/data/cognos/
products/bi_whitepaper.pdf
2. Heather Havenstein, “Better BI: Erlanger (Ky.) Police
Department,” Computerworld, September 1, 2008.
24
3. Gregg Keizer, “Mozilla wants to watch Firefox
users,” Computerworld, January 23, 2009.
4. Lisa A. Keefe, “Marketing Defined,” Marketing
News, January 15, 2007, pp. 28–29.
5. S. L. Cornish, “Product Innovation and the Spatial
Dynamics of Market Intelligence: Does Proximity
to Markets Matter?” Economic Geography, Vol. 73,
Iss. 2 (April 1997), pp. 147.
6. Michael Huster, “Marketing Intelligence: A First
Mover Advantage,” Competitive Intelligence Magazine, Washington: Vol. 8, Iss. 2 (March/April 2005),
p. 13.
7. http://www.wired.com/gadgets/wireless/
magazine/16-02/ff_iphone?currentPage=2.
8. Adapted and updated from John Kerrigan,
“Playing to Hispanics Garners Rewards,”
Marketing News, July 22, 2002, Vol. 36, Iss. 15, p. 20.
9. Pamela Parker, “The Population’s Building. Will
the Advertisers Come?” September 12, 2003,
http://www.clickz.com/experts/brand/buzz/
article. php/3076351.
10. Suzanne Vranica, “Miller Turns Eye Toward
Hispanics,” The Wall Street Journal—Market Place,
October 8, 2004, p. B. 3.
11. “His and Her Toothpaste,” Wall Street Journal,
September 5, 2002.
12. Adapted and updated from Ryan Nakashima,
“Turning Renters into Buyers,” The Hartford
Courant, June 28, 2005.
13. Adapted from P. Vincent Barbara and Gerald
Zaltman, Hearing the Voice of the Market. Cambridge,
MA: Harvard Business School Press, 1991.
14. George S. Day, “Evaluating Business Strategies,”
Strategic Planning Management, Spring 1984.
15. Douglas C. Bernhardt, “‘I Want It Fast, Actual,
Actionable’—Tailoring Competitive Intelligence to
Executive’s Needs,” Long Range Planning, Feb 1994,
Vol. 27, Iss. 1, p. 12.
16. B. Jaworski, L. G. Wee, “Competitive Intelligence
and Bottom-line Performance,” Competitive Intelligence Review, 1993, Vol. 3, No. 4, pp. 23–27.
17. “Key Role of Research in Agree’s Success Is Told,”
Marketing News, January 12, 1979, p. 14.
18. Dwight Silverman, “Compaq Unveils Revised Presarios,” Houston Chronicle, July 16, 1996, p. 1D.
19. Rohit Deshpande and Scott Jeffries, “Attitude
Affecting the Use of Marketing Research in Decision Making: An Empirical Investigation, in Educator’s Conference Proceedings, Series 47, Kenneth L.
Bernhardt et al. (eds.). Chicago: American
Marketing Association, 1981, pp. 1–4.
20. Rohit Deshpande and Gerald Zaltman, “Factors
Affecting the Use of Market Research Information:
A Path Analysis,” Journal of Marketing Research, 19,
February 1982, pp. 14–31; Rohit Deshpande and
Gerald Zaltman, “A Comparison of Factors Affecting Researcher and Manager Perceptions of Market
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Research Use,” Journal of Marketing Research, 21,
February 1982, pp. 32–38; Michael Hu, “An Experimental Study of Managers and Researchers Use of
Consumer Research, “Journal of Academy of Marketing Science, Fall 1986, pp. 44–51; and Hanjoon Lee,
Frank Acito, and Ralph Day, “Evaluation and Use
of Marketing Research by Decision Makers: A
Behavioural Simulation,” Journal of Marketing
Research, 24, May 1987, pp. 187–196.
Anil Menon and P. Rajan Varadarajan, “A Model
of Marketing Knowledge Use Within Firms,”
Journal of Marketing, 56, October 1992, p. 61.
Rohit Deshpande and Gerald Zaltman, “A Comparison of Factors Affecting Use of Marketing
Information in Consumer and Industrial Firms,”
Journal of Marketing Research, 24, February 1987,
pp. 114–118.
Betsy D. Gelb and Gabriel M. Gelb, “Coke’s Lesson
to the Rest of Us,” Sloan Management Review, Fall
1986.
Robert D. Hisrich and Michael P. Peters, Marketing
Decisions for New and Mature Products. New York:
Macmillan, 1991, p. 427.
Philip Cateora and John Graham (2007), “International Marketing,” 13th edition, McGraw-Hill
Irwin, pp. 533–534.
“Code of Ethics of Marketing Research Association,” Chicago, IL: Marketing Research Association, Inc.
“CASRO Code of Standards for Survey Research,”
The Council of American Survey Research Organizations, Annual Journal, 1992, pp. 19–22.
http://www.hoovers.com/
Michael W. Miller, “Coming Soon to Your Local
Video Store: Big Brother,” The Wall Street Journal,
December 26, 1990, p. 12.
“A Blockbuster of a Debate About Privacy: Blockbuster Entertainment May Sell Customer Information from Its Database on Film Rentals,” Washington
Post, January 1, 1991, p. E1.
Oswald A. J. Mascarenhas, “Exonerating Unethical Marketing Executive Behaviors: A Diagnostic
Framework,” Journal of Marketing, 59, April 1995,
p. 44.
“HEB Plans to Open Store in Monterrey,” Houston
Chronicle, September 10, 1996, p. C-1.
http://www.forbes.com/business/
walmartstores.com
James Flanigan, “Mexico Perplexes Cautious
Investors,” Houston Chronicle, September 25, 1999,
p. C-1.
V. Kumar, International Marketing Research (2000),
Prentice-Hall, NY.
25
CASE 1-1
Preteen Market—The Right Place
to Be In for Cell Phone Providers?
Do parents of preteens envision a need for equipping
their children with cell phones? This seems to be the
million-dollar question facing cell phone makers like
Firefly who view the preteen market as offering more
growth opportunities than the senior market. The preteen
market is unique in that the ultimate customer, the child,
does not make the purchase decision. On one side of this
indirect target market are parents with hectic or variable
schedules and those who are more harried or worried
about their children’s safety. On the other side are parents who are skeptical about the misuse of cell phones
by the kids themselves or others and those who consider
a kid cell phone as an unnecessary investment especially
considering the fact that kids are usually under some
supervision or the other, not to mention their tendency
to lose personal belongings.
Cell phone makers are following the trend set by
other industries such as food companies and retailers in targeting the preteen market. Firefly launched
its phones in February 2005 priced at $199 offering
12 months or 1,200 minutes of phone service—
whichever comes first. The Firefly phone has a number of safety features that make it simpler to use than
a conventional phone but harder to abuse. It does not
have a numerical keypad. Instead, parents can program up to 20 phone numbers into the device which
cannot be changed without a password. Parents can
also program the phone in such a way that it only
accepts calls from certain numbers.
Firefly is not the only wireless carrier targeting the
preteen market. Some providers offer regular cell
phones at a steep discount for family plans. Wherify, a
preteen phone specialist, is offering a kid phone that
comes with a global positioning system locator—so
parents can better track their offspring. Mattel Inc. is
launching a phone with a Barbie theme.
While equipping 8 to 12 year olds with cell phones
makes it easier for parents and children to get a hold
of each other and coordinate activities, parents have
other issues to grapple with. Many schools do not
allow students to bring cell phones to classrooms.
According to one University of Chicago child psychiatrist, most 8 to 12 year olds do not need a cell phone.
So the challenge for Firefly and their competitors is “At
what age do kids really need a phone?” and “Is the preteen market really worth going after?”
Questions for Discussion
Do you think preteens offer good market potential
for cell phone companies?
While designing products for this market, what are
the issues faced by the service providers?
Source: Adapted with permission from Mike
Hughlett, “Target: Teens—Cell Phone Company
Going After the Preteen Market,” Chicago Tribune,
April 12, 2005.
CASE 1-2
Best Buy on a Segmentation Spree
approach is Best Buy’s chief executive officer, Brad
Anderson. While most chains use their marketing
budgets primarily to maximize customer traffic in the
view that more visitors will lift revenue and profit,
Best Buy uses it to shun its unwanted customers. This
would mean dumping nearly 100 million of their
unprofitable customers. A variation of this approach
is being used in the financial-services industry,
wherein attention is lavished on the best customers
and the unprofitable customers are penalized with
fees for using ATMs or tellers or for obtaining bank
records.
From Best Buy’s 1.5 million daily customers, Mr.
Anderson wants to separate the “angels” from the
“devils.” Angels are customers who boost profits by
1.
2.
With $24.5 billion in 2005 sales, Best Buy Co. is the
nation’s largest seller of consumer electronics. Its spacious store and large inventory has helped it to increase
market share, even as its rivals such as Circuit City
Stores Inc. and Sears, Roebuck & Co., have struggled.
In fiscal 2004 that ended in February, Best Buy reported
net income of $570 million. Though up from $99 million during the 2003 fiscal period, it was still below the
$705 million it earned in fiscal 2002.
Best Buy estimates that nearly 20 percent of its
500 million visitors each year are undesirable and
wants to get rid of these customers. Behind this
26
snapping up high-definition televisions, portable electronics, and newly released DVDs without waiting for
markdowns or rebates. The devils are Best Buy’s worst
customers. They buy products, apply for rebates,
return the purchases, and then buy them back at
returned-merchandise discounts. They load up on
“loss leaders,” then flip the goods at a profit on eBay,
slap down rock-bottom price quotes from websites and
demand that Best Buy make good on its lowest-price
pledge. At dozens of websites like fatwallet.com, slickdeals.net, and techbargains.com, the devil customers
trade electronic coupons and tips from former clerks
and insiders, hoping to gain extra advantages against
the stores.
Best Buy has rolled out its angel–devil strategy in
about 100 of its 670 stores. It is examining sales
records and demographic data and devouring databases to identify good and bad customers. To lure the
high-spenders, it is stocking more merchandise and
more appealing service options. To keep the undesirables away, the company is cutting back on promotions
and sales tactics that tend to draw them. However, the
risk attached to such an exercise is significant. The
pilot stores have proved more costly to operate and
since different pilot stores target different types of
customers, they threaten the company’s economies
of scale.
After a series of analyses, Best Buy concluded that
most of its business came from five distinct customer
groups: upper-income men, also referred internally as
“Barrys”—who tend to be enthusiasts of action movies
and cameras, “Jills”—who are suburban mothers; busy
but usually willing to talk about helping their families,
small-business owners, “Buzzes”—technology enthusiasts, also known as early adopters and interested in
buying and showing off the latest gadgets. Mr. Anderson
decided that each store should study the demographics of the local market and realign themselves by
stocking merchandise accordingly. Armed with this
information, Best Buy began working on ways to avoid
unprofitable customers. They just could not bar them
from entering their stores. But in summer 2004, they
took steps to put an end to one of the biggest problems.
They enforced a restocking fee of 15 percent of the purchase price on returned merchandise. To discourage
customers who return items with the intention of
repurchasing them at an “open box” discount, they are
experimenting with reselling them over the Internet, so
that the goods do not reappear in the stores where they
were originally purchased. In July 2004, they cut ties
with fatwallet.com, an online affiliate that collected
referral fees for delivering customers to Best Buy’s
website, for revealing information about Best Buy’s
planned Thanksgiving Weekend sale.
Training of store clerks plays a vital role in identifying desirable customers by concentrating on their
shopping preferences and behavior. Staffers use quick
interviews to identify profitable customers. For
instance, if a customer says his family has a regular
“movie night,” he is a prime candidate for home-theater
equipment. Likewise, shoppers with large families are
shown larger appliances and time-saving products.
The company attempts to entice the Barrys and Jills by
providing services like a “personal shopper,” to locate
unusual items, alert them on preferred items, and coordinate service calls.
Best Buy’s Westminster, California, store is one of
100 using this technique. It targets upper-income men
with a wide range of pricey home-theater systems and
small-business owners with network servers and office
PC solutions. During DVD releases, the store clerks
identify promising customers and guide them into a
back room that displays $12,000 high-definition hometheater systems. The room has easy chairs, a leather
couch, and a basket of popcorn to simulate a theater
atmosphere. At stores popular with Buzzes, Best Buy is
setting up video game sections furnished with leather
chairs and players hooked to giant plasma-screen televisions. Mr. Anderson says early results indicate pilot
stores performed far better than conventional stores,
following which the company began converting
another 70 stores.
Best Buy intends to customize the rest of its stores
over the next three years. With customization of stores,
stock maintenance becomes crucial as it could topple
the sales and customer goodwill. Periodic design
changes and customizations cause the costs to go up,
but Mr. Anderson says that as stores share the successful ideas for acquiring customers, the average cost per
store should fall.
Questions for Discussion
1.
2.
What benefits can Best
customer segmentation
Do you think Best Buy
marketing intelligence
decisions?
Buy hope to gain from its
tactics?
is making adequate use of
practices for its business
Source: Adapted with permission from
Gary McWilliams, “Analyzing Customers,
Best Buy Decides Not All Are Welcome,”
The Wall Street Journal, November 8, 2004, p. A.1.
27
CASE 1-3
Ethical Dilemmas in Marketing
Research1
What action would you take?
4.
The following scenarios present a set of ethical dilemmas that might arise in marketing research. Your
assignment is to decide what action to take in each
instance. You should be prepared to justify your decision. Bear in mind that there are no uniquely right
answers: Reasonable people may choose different
courses of action.
A well-respected public figure is going to face trial
on a charge of failing to report his part ownership of
certain regulated companies while serving as a Canadian provincial minister. The defense lawyers have
asked you, as a market research specialist, to do a
research study to determine the characteristics of
people most likely to sympathize with the defendant
and hence to vote for acquittal. The defense lawyers
have read newspaper accounts of how this approach
has been used in a number of instances.
What action would you take?
1.
You are the market research director of a pharmaceutical company, and the executive director
suggests to you that company interviewers telephone physicians under the name of a fictitious
market research agency. The purpose of the
survey is to help assess the perceived quality
of the company’s products, and it is felt that the
suggested procedure will result in more objective
responses.
5.
What action would you take?
2.
You are employed by a marketing research firm
and have conducted an attitude study for a client.
Your findings indicate that the product’s marketing efforts are not effective. This finding is badly
received by the client’s product management team.
They request that you omit that data from your
formal report, which you know will be widely distributed, on the grounds that the oral presentation
was adequate for their needs.
What do you do?
3.
You are a study director for a research company
undertaking a project for a regular client of your
company. A study you are working on is about to
go into the field when the questionnaire you sent
to the client for final approval comes back drastically modified. The client has rewritten it, introducing leading questions and biased scales. An
accompanying letter indicates that the questionnaire must be sent out as revised. You do not
believe that valid information can be gathered
using the revised instrument.
You are the market research director for a large
chemical company. Recent research indicates that
many of your company’s customers are misusing
one of its principal products. There is no danger
resulting from this misuse, though customers are
wasting money by using too much of the product
at one time. You are shown the new advertising
campaign by the advertising agency. The ads not
only ignore this problem of misuse, they actually
seem to encourage it.
What action would you take?
6.
You show up your first day for a summer internship to meet your supervisor and get your first
assignment. She gives you a questionnaire with
specific marketing planning questions and tells
you that she would like you to contact the
company’s main three competitors and tell them
you are a student doing a study on the industry
and get the answers to the questions. She says
you should not tell them you are working for the
company over the summer and adds “there is
nothing wrong with not telling them—omitting
the fact is not really lying.” You ask her what you
should do if they ask, and she says ‘just
tell them you are a student working on a paper
and you don’t want to get a poor grade on the
assignment.”
What action would you take?
1
These vignettes were provided through the courtesy of Professor Charles Weinberg, University of British Columbia, and are reproduced with his
permission.
28
MARKETING RESEARCH
IN PRACTICE
2
Learning Objectives
■ Discuss briefly the practice of marketing research.
■ Understand the concept of information systems and decision support systems.
■ Explain marketing decision support systems.
■ Introduce the various suppliers of marketing research information and the types of services
offered by them.
■ Talk briefly about the criteria used to select suppliers.
■ Introduce the career options available in the marketing research industry.
In practice, a marketing research department’s goal can be grouped into three major categories:1
programmatic, selective, or evaluative. Programmatic research is performed to develop marketing options through market segmentation, market opportunity analysis, or consumer attitude and
product usage studies. Selective research is done to test different decision alternatives such as
new product concept testing, advertising copy testing, pretest marketing, and test marketing.
Evaluative research is carried out to evaluate performance of programs, including tracking advertising recall, corporate and brand image studies, and measuring customer satisfaction with the
quality of the product and service. As the number of products and types of services introduced
into the market increase, the need for marketing research explodes and the future of marketing
research appears to be both promising and challenging.2
Unquestionably, marketing research is a growth industry. In the last decade, real expenditures on marketing research (that is, after adjusting for inflation) more than doubled. This is
largely a consequence of economic and social changes that have made better marketing an
imperative.
With marketing the new priority, marketing research is the rallying cry. Companies are frantically trying to get their hands on information that identifies and explains the needs of powerful new consumer segments now being formed. Kroger Co., for example, holds more than 250,000
consumer interviews a year to define consumer wants more precisely. Some companies are pinning their futures on product innovations, others are rejuvenating time-worn but proven brands,
and still others are doing both.3
Not only are the companies that always did marketing research doing a great deal more, the
breadth of research activities also continues to expand.
■ Senior management is looking for more support for its strategic decisions; therefore, researchers
are doing more acquisition and competitor studies, segmentation and market structure analyses,
and basic strategic position assessments.
■ Other functions, such as the legal department, now use marketing research evidence routinely.
Corporate Affairs wants to know shareholders’, bankers’, analysts’, and employees’ attitudes
toward the company. The service department continuously audits service delivery capability
and customer satisfaction.
29
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
■ Entire industries that used to be protected from the vagaries of competition and changing
customer needs by regulatory statutes are learning to cope with a deregulated environment.
Airlines, banks, and financial-services groups are looking for ways to overcome product proliferation, advertising clutter, and high marketing costs brought on by more sophisticated
customers and aggressive competitors.
In this chapter we will look at how companies use the information gathered by marketing
research, at the various ways they obtain this marketing research information, and at the career
opportunities available in the marketing research industry.
INFORMATION SYSTEMS, DECISION SUPPORT
SYSTEMS, AND MARKETING RESEARCH
An information system (IS) is a continuing and interacting structure of people, equipment, and
procedures designed to gather, sort, analyze, evaluate, and distribute pertinent, timely, and accurate information to decision makers. While marketing research is concerned mainly with the
actual content of the information and how it is to be generated, the information system is concerned with managing the flow of data from many different projects and secondary sources to
the managers who will use it. This requires databases to organize and store the information and
a decision support system (DSS) to retrieve data, transform it into usable information, and disseminate it to users.
Databases
Information systems contain three types of information. The first is recurring day-to-day information, for example, the market and accounting data that flow into the organization as a result
of market analysis research and accounting activities. Automobile firms use government sources
for monthly data on new-car sales by brand and geographic area. In addition, surveys are conducted yearly to determine the ages and types of automobiles currently being driven, the
lifestyles of the drivers (their activity and interest patterns), their media habits, and their intentions to replace their cars. The accounting department submits sales and inventory data for each
of its dealers on a continuing basis, to update and supplement the information system.
A second type of information is intelligence relevant to the future strategy of the business.
Automobile firms, for example, collect reports about new sources of fuel to power automobiles.
This information might come from scientific meetings, trade organizations, or perhaps from government reports. It also might include information from salespersons or dealers about newproduct tests being conducted by competitive firms. Intelligence is difficult to develop, because
it usually involves diverse and changing sets of topic and information sources and is rarely collected systematically.
A third input to the information system is research studies that are not of a recurring nature.
The potential usefulness of a marketing research study can be multiplied manifold if the information is accessible instead of filed and forgotten. However, the potential exists that others may
use the study, although perhaps not in the way it was originally intended.
Decision Support Systems
30
Databases have no value if the insights they contain cannot be retrieved. A decision support system not only allows the manager to interact directly with the database to retrieve what is wanted,
it also provides a modeling function to help make sense of what has been retrieved.
A common example of a DSS in action is that used by many industrial salespeople—especially
those selling products that require significant customization. The salesperson frequently will be
INFORMATION SYSTEMS, DECISION SUPPORT SYSTEMS,
AND
MARKETING RESEARCH
asked whether or not the price and delivery time of a unique product configuration will meet or
exceed a competitor’s promises. Without leaving the customer’s office, the salesperson can plug
a laptop computer into a phone jack and begin communicating with a database stored in the company’s main computer memory. The salesperson types in the product configuration and desired
delivery data, and these requirements are compared to the costs, inventory, and assembly time
contained in the databank. In a matter of minutes, the salesperson can propose a price and delivery date—and perhaps close the sale.
Each firm has to develop or adapt models to support its own decision problems. For example, Avon Products, Inc., the door-to-door cosmetics firm, has unique problems as the result of
a part-time sales force of almost 400,000 representatives theoretically covering half of the 80 million households in North America. This sales force carries a large product line that each year
adds 1,600 new or reformulated products. The following computer models were added to their
DSS to help cope with these problems:4
■ A sales force turnover model, which revealed that the most significant variable influencing
the turnover rate was the level of the appointment fee that representatives pay for initial
materials.
■ An order model that explains the components of the average order and isolates the actionable
variables, such as the size and timing of the catalog and the gift incentives.
■ A procurement model that helps determine how much of a new product to buy, when to purchase it, and the risks involved.
Applying Information Systems to Marketing Research
Often the process of developing and using models and information systems reveals gaps in the
databank that have to be closed. These emergent needs for information become a marketing
research problem; for example,
■ Performance (sales, market share, contributions, patronage) may be unsatisfactory relative to
objectives. Perhaps the condition can be traced to a specific geographic area, but the underlying reasons still must be sought before action can be taken.
■ A competitor may launch a new product or employ a new advertising appeal, with unknown
consequences for the firm’s competitive position.
■ An unavoidable increase in costs puts pressures on profitability (or, in the case of a transit system, for example, increases subsidy requirements to an unacceptable level). Various possible
increases in prices (or fares) must be evaluated.
■ An upsurge in interest in health and nutrition may suggest to a snack company a new product line directed toward responding to this interest. Concept testing might be a first step in
exploring this opportunity.
Given the sometimes chaotic and usually uncertain nature of most market environments, a
large number of problems and some opportunities can emerge. Few will ever be given formal
consideration. There may be no further need for clarification, the implications may not appear
serious, or the response may appear evident in the judgment of the decision maker. Our interest
is in those problems or opportunities that need to be clarified, whose consequences are uncertain,
or that involve the development of new programs, products, or services.
The information system serves to emphasize that marketing research should not exist in isolation as a single effort to obtain information. Rather, it should be part of a systematic and continuous effort by the organization to improve the decision-making process. Marketing Research
in Action 2-1 describes the “fourth wave” in marketing research, telling how marketing research
has progressed from the rudimentary to the sophisticated.
31
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
MARKETING RESEARCH IN ACTION 2-1
The Fourth Wave in Marketing Research
Much of the initial interest in marketing research came about during the transition from a sales-oriented
to a marketing-oriented business environment. In the First Wave, seat-of-the-pants decision making progressed to a databased decision. But, with more data available, a large problem arose: A handful of data
was helpful; a truckload was not necessarily better. So, in the Second Wave, the progression was from
databased decision to information-based decision making. Rather than review a multitude of individual
facts, the role of marketing research evolved to manipulate data to summarize the underlying patterns.
Unfortunately, much of the criticism of marketing research today is a result of this excessive focus on
methodology and statistics. The problem now with marketing research was not centered on incompetent analysis, but the lack of a decision-maker’s perspective and the inability to provide actionable
insights consistently once the data were analyzed correctly.
In the Third Wave, we progressed from information-based decision making to system-based decision making. The Third Wave involved a number of developments centered on automated decision
systems (ADS) that put the power of marketing information directly into the hands of nontechnical
decision makers. It was a three-way marriage between marketing analysis, computer and information
technologies, and the formal marketing planning process.
The team approach, including marketing researchers, systems engineers, and marketing management,
provided the synergy needed to build useful computer-based systems that helped managers through the
marketing planning and evaluation process. Through pooling expertise and applying information technologies, the full potential of marketing information was realized in the Third Wave.
Finally, in the Fourth Wave, the emergence of e-commerce and large customer databases has motivated organizations to move from an aggregate-level summarization of data (in order to uncover the
underlying patterns as in the Second Wave) to a manipulation of data on an individual-by-individual
basis. A majority of organizations currently are leveraging the technology developed during the Third
Wave and the rich information they have about each individual customer, to tailor both their product/
service and marketing messages on an individual customer basis. Also called mass-customization, this
strategy develops relations with every customer that an organization possesses and finally allows
managers to assess the lifetime value of every individual customer to the organization.
SOURCE: Adapted from Kumar, Aaker, and Day, “Essentials of Marketing Research,” 2nd ed. [2002], John Wiley & Sons, Inc.; and “Third Wave of Marketing
Research on the Horizon,” Marketing News [March 1, 1993], p. 6.
MARKETING DECISION SUPPORT SYSTEMS
32
A typical marketing manager regularly receives some or all of the following data: factory shipments or orders; syndicated aggregate (industry) data services; sales reports from the field
sales force; consumer panel data; scanner data; demographic data; and internal cost and
budget data. These data may also come in various levels of detail and aggregation. Often they
use different reporting periods and incompatible computer languages. Add to this sales estimates about competing brands and advertising, promotion, and pricing activity, and there is
a data explosion.
Managers don’t want data. They want, and need, decision-relevant information in accessible
and preferably graphical form for (1) routine comparisons of current performance against past
trends on each of the key measures of effectiveness, (2) periodic exception reports to assess which
sales territories or accounts have not matched previous years’ purchases, and (3) special analyses to evaluate the sales impact of particular marketing programs, and to predict what would
happen if changes were made. In addition, different divisions would like to be linked to enable
product managers, sales planners, market researchers, financial analysts, and production schedulers to share information.
The purpose of a marketing decision support system (MDSS) is to combine marketing data
from diverse sources into a single database which line managers can enter interactively to
quickly identify problems and obtain standard, periodic reports, as well as answers to analytical questions.
MARKETING DECISION SUPPORT SYSTEMS
Manager
Modeling
Display
Analysis
Database
Environment
FIGURE 2-1
The four components of a MDSS.
Characteristics of a MDSS
A good MDSS should have the following characteristics.
1. Interactive: The process of interaction with the MDSS should be simple and direct. With just
a few commands the user should be able to obtain the results immediately. There should be
no need for a programmer in between.
2. Flexible: A good MDSS should be flexible. It should be able to present the available data in
either discrete or aggregate form. It should satisfy the information needs of the managers in
different hierarchical levels and functions.
3. Discovery oriented: The MDSS should not only assist managers in solving the existing problems but should also help them to probe for trends and ask new questions. The managers
should be able to discover new patterns and be able to act on them using the MDSS.
4. User friendly: The MDSS should be user friendly. It should be easy for the managers to learn
and use the system. It should not take hours just to figure out what is going on. Most MDSS
packages are menu driven and are easy to operate.
A typical MDSS is assembled from four components (see Figure 2-1).
Database
The database contains data from all sources, stored in a sufficiently disaggregated way so that
it can be analyzed by product item, sales district, trade account, or time period. The best systems have databases that can be easily updated with new information and have sufficient flexibility that data can be readily analyzed in new ways. Since most analyses deal with a subset of
a larger database, the supporting software should permit random access to any and all data to
create appropriate subsets.
Reports and Displays
The capabilities of a MDSS range from simple ad-hoc tables and reports to complex plots, charts,
and other graphic displays. Any report or display can include calculations such as variances and
running totals, or the results of statistical procedures found in the system. Typical reports produced
with a MDSS include status reports that track current trends, exception reports on troubled brands
and markets, and variance reports showing budget and actuals for sales and profits. The report in
Figure 2-2 answers the question: “In which regions is the Mama Mia brand not keeping up with
the industry trend?” Figure 2-2 shows that product category sales were up 5.8 percent, whereas
Mama Mia sales were up only 2.0 percent. On a more detailed level, product category sales were
33
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
Comparison Matrix
Issue 203 Ending April 2, 2003
% Change in Category Sales vs One Year Ago
28.7
20.3
13.3
13.1
12.8
12.8
11.6
10.9
10.1
9.6
9.3
9.3
9.3
8.5
8.5
8.4
7.9
7.8
7.8
7.8
7.1
7.1
7.0
6.6
6.0
5.9
Above
U.S. Average
Below
% Change in Mia Sales vs One Year Ago
Memphis
San Antonio
Oklahoma City
Cleveland
Kansas City
Detroit
Cincinnati/Dayton
Indianapolis
9.8
14.0
10.5
6.5
9.3
4.1
4.1
0.0
Houston
Milwaukee
Norfolk/Richmond
6.3
⫺0.4
0.9
Phoenix/Tucson
Minneapolis/St. Paul
Omaha/Des Moines
St. Louis
6.4
8.4
6.3
12.1
⫺1.5
Pittsburgh
Buffalo
New Orleans
Albany/Schenectady
7.1
7.7
⫺4.5
New York
Salt Lake City
Boston/Providence
Dallas/Ft. Worth
Jacksonville
Syracuse
3.1
11.1
7.7
6.7
7.8
1.3
Charlotte
4.0
Portland
Nashville/Knoxville
3.6
3.2
Denver
2.6
5.8
5.3
4.8
4.6
4.4
3.9
2.9
2.7
1.7
1.2
0.3
⫺0.6
⫺2.3
⫺5.2
Birmingham/Mobile
⫺0.9
Baltimore/Washington
⫺3.1
Seattle/Tacoma
Atlanta
Miami
Chicago
Los Angeles/San Diego
Philadelphia
Raleigh/Greensboro
San Francisco
Below
7.1
10.1
⫺0.3
1.3
⫺0.2
⫺4.7
⫺0.2
⫺9.6
2.0
U.S. Average
Above
FIGURE 2-2
Comparison of brand performance with industry trend.
up 13.1 percent in Cleveland and 1.7 percent in Miami, while Mama Mia sales rose 6.5 percent in
Cleveland and fell 0.3 percent in Miami.
Analysis Capabilities
34
Analysis capabilities are used to relate the data to the models, as well as to clarify relationships,
identify exceptions, and suggest courses of action. These capabilities should include the ability
to make calculations such as averages, lags, and percentage changes versus a previous period,
and to conduct seasonal analyses and standard statistical procedures such as regression, correlation, and factor analysis. These procedures will be covered in subsequent chapters of this book.
MARKETING DECISION SUPPORT SYSTEMS
Models
Models represent assumptions about how the world works, and specifically how brand sales,
shares, and profits respond to changes in elements of the marketing mix. Models are used to test
alternative marketing programs, answer “what if” questions, and assist in setting more realistic
objectives. For example, managers want help with such questions as: What is the impact on profitability of achieving wider distribution? What is the optimal call level for each sales representative for each account and prospect? What objectives should be set for coupon redemption and the
profitability of promotion programs? The models used to address these questions can range from
forecasts to complex simulations of relationships among marketing, economic, and other factors.
Using a MDSS offers the organization a number of advantages. It results in substantial cost
savings, because it helps in making better and quicker decisions. The presence of a MDSS forces
the decision maker to view the decision and information environment within which he or she operates; hence, it leads to a better understanding of the decision environment. Since managers can now
retrieve and utilize information that was never accessible before, it results in the enhancement of
decision-making effectiveness. Using a MDSS results in better quality and quantity of data being
collected and hence increases the value of the information to managers.
Gaining Insights from a MDSS
When an over-the-counter (OTC) drug manufacturer suffered a decline in national unit market share
for its drug “Alpha,” management turned to a MDSS for insights. They suspected the losses would
be traced to actions of the two main competitors. “Beta” was a private-label competitor that was sold
at half the price of Alpha. The other competing brand, “Delta,” was produced and marketed by
another division of the same company, following a rather similar strategy. Initial data from the decision support system seemed to confirm management’s initial suspicions. Alpha’s share had dropped
from 5.0 percent to 2.5 percent, and Delta’s share had more than doubled, from 2.0 percent to 4.5 percent. However, subsequent analysis of the database showed that this information was misleading.
The premise of the further analysis was that any competitive effects should be evident at the
regional as well as the national level. To test this possibility, the market share changes of Alpha
were related to share changes of Beta and Delta, by region, for a six-month period.
The results, shown in Figure 2-3, confirmed the adverse effects of Beta on Alpha. In almost
every region, a share decrease for Alpha was associated with a share increase for Beta. A different
Change in
Alpha’s
Share in
Share
Points
0.5
Area 6
0.3
Area 3
0.0
Area 5
Area 1
ⴚ0.3
Area 2
Area 9
ⴚ0.5
ⴚ0.8
ⴚ1.0
ⴚ1.3
Area 7
Area 8
Area 4
ⴚ1.5
ⴚ1.0
0.0
1.0
2.0
Change in Beta’s
Share in Share Points
FIGURE 2-3
Regional market share changes—Alpha and Beta.
3.0
4.0
35
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
Change in
Alpha’s
Share in
Share
Points
0.50
0.25
Area 2
0.00
Area 5
Area 6
ⴚ0.25
Area 1
ⴚ0.50
Area 7
Area 9
Area 3
Area 8
ⴚ0.75
ⴚ1.00
Area 4
ⴚ1.25
0.0 0.5 1.0
1.5
2.0
2.5
Change in Delta’s
Share in Share
Points
FIGURE 2-4
Regional market share changes—Alpha and Delta.
3.0
3.5
picture emerged, however, when a similar analysis (shown in Figure 2-4) was done with Delta.
In the regions where Alpha’s share had decreased the least, Delta’s share had increased the most.
Conversely, Alpha’s share had decreased the most in those regions where Delta had gained the
least share. Clearly, Delta was not the source of Alpha’s problems; more likely, Delta was helping Alpha by combining the two brands’ sales force efforts. This analysis prevented a potentially
damaging interdivisional dispute and helped focus management’s attention on the proper target
competitor.
SUPPLIERS OF INFORMATION
In general, managers can acquire the necessary information for decision making from two basic
types of sources:
1. The corporate or in-house marketing research department
2. External suppliers
Usually, managers use a mix of in-house and external approaches to solve a certain problem.
Both can feed information directly to their clients, who are users with decision-making needs.
More often, the outside suppliers get their direction and provide information to an inside
research group. These inside suppliers translate the problems of their clients into specific information requirements, decide how the information will be collected and by whom, and then interpret the findings. Figure 2-5 shows the interaction among the participants in a marketing research
activity.
The purpose of this section is to discuss briefly the nature and attributes of the providers of
marketing research services, the types of services they provide, and the factors that influence the
choice of a suitable supplier for a given situation. Figure 2-6 gives a concise summary of the different types of information suppliers within the marketing research industry.
Corporate/In-House Marketing Research
36
The location of the marketing research department within an organization and the strength of
the department vary from firm to firm and to a very great extent depend on the requirements
SUPPLIERS
OF INFORMATION
Information Users
•
•
•
•
•
General management
Planning
Marketing and sales managers
Product managers
Lawyers
Information Suppliers:
In Company
•
•
•
•
Information Suppliers:
Outside Company
Marketing research department
Sales analysis group
Accounting department
Corporate strategic planning
•
•
•
Research consultants
Marketing research suppliers
Advertising agencies
FIGURE 2-5
Participants in marketing research activities.
for information and the organizational structure of the firm. Some firms have a single centralized research department, housed in the corporate headquarters, which provides the information required to the various business units scattered geographically and/or functionally. The
other extreme is the completely decentralized operation, wherein each business unit or geographic unit has its own research department. The type of structure adopted depends on the
amount of information required, the frequency with which it is required, the uniqueness of the
information, and the time available to collect it. In most major organizations, especially in multinational corporations, a mix of both these structures can be found.
Information
Supplier
Corporate
In-house Supplier
Structured
(independent
department)
External
Supplier
Unstructured
(one or more
MR employees)
Customized
Services
Standardized
Services
Syndicated
Services
FIGURE 2-6
Information suppliers and services.
Field
Services
Branded
Product Services
Selective
Services
37
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
Not all organizations (regardless of size) have an in-house research establishment. Even
among those that have an in-house research department, it is not unusual to seek the assistance
of external suppliers. Virtually all research users at some time use the services of outside research
specialists. Their role may be limited to raw-data collection, depending on the research approach,
questionnaire, and sampling method provided by the client. At the other extreme, the client may
assign the entire problem to an outside consultant who is responsible for every step to the completed report and action recommendations. Other possibilities are to bring in outside specialists
for special problems (such as a sampling expert to draw a complex sample), or to employ services that have special facilities or data.5
Many related considerations influence the decision to go outside:
1. Internal personnel may not have the skills or experience. Few but the largest companies
have specialists in all areas, from psychologists able to conduct focus group interviewing
to electronics engineers with MBAs who have studied the telecommunications equipment
market.
2. Outside help may be called in to boost internal capacity in response to an urgent deadline.
3. It may be cheaper to go outside. Specialists who have encountered similar problems with other
clients probably are more efficient in dealing with the problem, and because they are not on
the staff there is no risk of underutilization of their time.
4. Shared cost and multiclient studies coordinated by an outside supplier offer considerable
savings possibilities. Multiclient studies are feasible when several organizations have
related needs for information about a major topic, such as the future of electronic funds
transfer systems. Each client pays an agreed share of the total cost. The ultimate in sharedcost studies are the large standardized data collection services, such as store audits of product sales activity or omnibus surveys, which combine questions from several clients.
5. Often, outside suppliers have special facilities or competencies (an established national interviewing field force, conference rooms with one-way mirrors, banks of telephone lines, or test
kitchens) that would be costly to duplicate for a single study.
6. Political considerations may dictate the use of an outside research specialist whose credentials are acceptable to all parties in an internal policy dispute. Research people within
the organization may be well advised to avoid being on one side or the other of a sensitive issue.
7. Marketing research is used increasingly in litigation or in proceedings before regulatory or
legislative bodies. The credibility of the findings generally will be enhanced if the study is
conducted by a respected outsider. Also, this kind of research often is subjected to critical
questioning or cross-examination and is likely to stand up only if designed to high standards, which may exceed those used within the organization for routine decision-making
purposes.
External Suppliers of the Research Industry
The marketing research industry consists of several hundred research firms, ranging from
small, one-person operations to large corporations having operations in multiple countries.
Table 2-1 lists the top 25 U.S. research firms and their annual revenues for the year 2008. The
table also gives an estimate of the total revenue from foreign operations.
According to Jack Honomichl, the president of Marketing Aid Center, Inc. (Barrington, IL)
and the author of the Annual Industry Review, the U.S. marketing research industry is becoming
more worldly and cosmopolitan. As can be seen from the table, the total revenue of the top 25
firms exceeds $21 billion, and more than one half of this revenue is estimated to come from foreign operations.6 The total revenue in 2008 represents a 1.6 percent growth over 2007. After
adjusting for worldwide inflation, the real growth was ⫺2.2 percent, indicating a downturn in
more than 20 years.
38
SUPPLIERS
OF INFORMATION
TABLE 2-1
Top 25 U.S. Research Organizations
U.S. Ranks
2010 2009 Organization
Headquarters
1
1
The Nielsen Co.
New York
2
2
Kantar*
London & Fairfield, Conn.
Global
Research
Revenue ($,
in millions)
Revenues From
Outside Parent
Country ($,
in millions)
$4,958.0
$2,551.0
3,183.6
2,387.7
Percent of Global
Revenues from
Outside Home
Country (%)
51.5%
75.0
3
3
IMS Health Inc.*
Norwalk, Conn.
2,211.6
1,410.5
63.8
4
4
GfK SE
Nuremberg
1,716.2
1,265.4
73.7
5
5
Ipsos SA
Paris
1,512.8
1,383.0
91.4
6
6
Synovate
London
884.8
776.3
87.7
7
7
SymphonylRI Group
Chicago
727.0
270.0
37.1
8
8
Westat Inc.
Rockville, Md.
455.3
—
9
10
INTAGE Inc.**
Tokyo
416.2
12.7
3.1
10
9
Arbitron Inc.
Columbia, Md.
395.4
5.0
1.3
—
11
12
The NPD Group Inc.
Port Washington, N.Y.
240.1
66.4
27.7
12
11
J.D. Power and
Associates*
Westlake Village, Calif.
231.4
84.1
36.3
13
14
Video Research Ltd.** Tokyo
212.0
—
—
14
17
IBOPE Group
210.8
38.8
18.4
15
—
ICF International Inc. Fairfax, Va.
200.1
46.9
23.4
16
18
comScore Inc.
175.0
32.7
18.7
17
15
Harris Interactive Inc. New York
166.8
72.9
43.7
18
16
Maritz Research
Fenton, Mo.
162.1
21.2
13.1
19
20
Abt SRBI Inc.
Cambridge, Mass.
129.4
42.6
32.9
20
25
Macromill Inc.
Tokyo
115.0
—
—
21
22
Lieberman Research
Worldwide
Los Angeles
100.8
18.5
18.4
22
19
ORC International
Princeton, N.J.
98.0
34.1
34.8
23
21
Mediametrie
Paris
92.3
8.5
9.2
24
23
Cello Research &
Consulting
London
84.5
37.2
44.0
25
24
Market Strategies
International
Livonia, Mich.
80.7
13.1
16.2
$18,759.9
$10,578.6
Sao Paulo
Reston, Va.
Total
56.4%
SOURCE: Jack Honomichl, “Top 25 Report,” Marketing News, August 30, 2011: p. 16.
Regarding the performance of U.S. marketing research industry, the spending for marketing
research services grew 1.5 percent in 2008, compared with 2007.7 Figure 2-7 provides the growth
in research spending within U.S. between 1988 and 2008. While it is rather alarming to note that
the least growth occurred in the most recently reported year, 2008; it may not come as a surprise
given the difficult economic environment.
Table 2.1 shows that in 2008, at least 20 out of the Top 25 firms reported income from outside the United States—typically through a branch office. For instance, $10.6 billion or 54 percent of the Top 50 firms’ total revenue came from outside the U.S. Therefore, it is important to
understand that while focus on U.S. performance is necessary, cognizance of the global research
industry potential is essential. The performance of the Global Top 10 firms in 2007 is covered in
The International Marketing Research Industry section of this chapter.
39
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
14.0
12.6%
12.0
11.6%
11.5%
10.1%
10.0
9.2%
9.5%
9.2%
9.9%
9.0%
8.6%
8.0
6.6%
Percentage Growth
6.0
5.7%
5.6%
4.0%
4.0
6.0%
4.5%
3.5%
1.5%
2.0
0.0
5.5%
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
−0.0
−0.2
−0.4
−4%
Year
FIGURE 2-7
Growth in Research Spending within the United States.
SOURCE: Jack Honomichl, “Top 50 Report,” Marketing News, June 30, 2011: p. 13.
Type and Nature of Services
External suppliers that collectively comprise the marketing research industry can be further classified into six different groups, depending on the type and nature of the services they provide. Based
on Figure 2-6, the major types of services provided by external suppliers of information are discussed in the paragraphs that follow.
Customized Services. Firms that specialize in customized services work with individual clients to
help them develop and implement a marketing research project from top to bottom. They work with
management on any given problem and go through the entire research process, including analyzing
and presenting the final results.
Syndicated Services. Syndicated services are companies that routinely collect information on
several different issues and provide it to firms that subscribe to their services. The Nielsen Television Index, which provides information on audiences viewing different TV programs, is an
example of such services. These suppliers also provide information on retail sales, household
purchasing patterns, and so on, which they collect through scanner data.
Standardized Services. Standardized services are market research projects conducted in a standard, prespecified manner and supplied to several different clients. The Starch Readership Survey is a typical example of such a service; it provides clients information regarding the effectiveness of print advertisements, in the form of Starch scores.
40
Field Services. Field services suppliers concentrate on collecting data for research projects. They
specialize in various survey techniques such as mail surveys, telephone surveys, or personal surveys. These organizations range from small, one-person establishments to large, multinational,
wide-area telephone service (WATS)-line interviewing services, and have extensive facilities for
personal interviews in homes and shopping malls. Some of these firms specialize in qualitative
data collection methods such as focus group interviews and projection techniques.
THE INTERNATIONAL MARKETING RESEARCH INDUSTRY
Selective Services. Some companies specialize in just one or two aspects of marketing research,
mainly concerned with data coding, data editing, or data analysis. These generally are small
firms, sometimes referred to as “lab houses,” with expertise in sophisticated data analysis techniques. The proliferation of software for marketing research projects has led to an increase in
such lab houses. Any enterprising individual with expertise in computer and sophisticated multivariate analysis techniques can acquire the necessary software packages and establish a firm
specializing in data analysis.
Branded Products Services. Some firms have developed specialized data collection and analysis procedures to address specific types of research problems, which they market as branded
products. PRIZM, a Claritas Corporation product that forms clusters of the population on the
basis of lifestyle and Zip Code classifications, is an example of one such branded product.
Using this technique, the entire U.S. population is divided into 40 clusters. Author Michael J.
Weiss, in his popular book, Clustering of America, New York, NY: Harper & Row Publishers
(1988), provides a vivid portrait of the nation’s 40 neighborhood types—their values, lifestyles,
and eccentricities.
CRITERIA FOR SELECTING EXTERNAL
SUPPLIERS
Once the decision has been made to go outside, there remains the question of which consultant
or supplier to retain. What criteria should a firm adopt in selecting an external research supplier?
Several academic scholars have conducted research studies to identify the factors that are important in the selection of external suppliers. A crucial factor in the choice is the judgment as to
whether the supplier or consultant actually can deliver the promised data, advice, or conclusions.
This judgment should be made only after the following steps have been followed:
1. A thorough search for names of people and companies who have acknowledged expertise in
the area of the study8
2. Selection of a small number of bidders on the basis of recommendations of colleagues or others who have had similar needs
3. Personal interviews with the person who would be responsible for the project, asking for
examples of work on similar problems, their procedures for working with clients, and the
names of previous clients who could provide references
4. A check of the references of each potential supplier, with special attention to comments on
their depth of competence and expertise, their creativity in dealing with problems, and the
quality and adequacy of resources available
Selection is made on the basis of how well the problem and objectives have been understood,9 comments by the references, and whether the quoted price or fee is a good value in
light of the research approach that is proposed. Seldom is the lowest quotation the best value.
To minimize the problem of comparability, all bidders should respond to the same study
specifications.10
THE INTERNATIONAL MARKETING
RESEARCH INDUSTRY
The growing importance of the international components of the marketing research industry revenue reflects the fact that the United States accounts for over 60 percent of the combined global
research revenue of the top 25 research organizations. United Kingdom accounts for 18 percent
of research revenues, Japan accounts for 3 percent, and the remainder is accounted for by the rest
of the world. Almost all major countries in the world, including the newly industrialized and
41
CHAPTER 2 MARKETING RESEARCH
IN
PRACTICE
TABLE 2-2
Top 25 Global Research Organizations
Organization
Country
No. of Countries
with Subsidiaries/
Branch Offices
Research only
Full-time
Employees
Global
Research
Revenues
(millions)
Revenues from
Outside
Country, US$
(millions)
The Nielsen Co.
U.S.
100
32,900
$4,958.0
$2,551.0
Kantar*
U.K.
80
21,800
3,183.6
2,387.7
IMS Health Inc.*
U.S.
74
7,000
2,211.6
1,410.5
GfK SE
Germany
65
10,546
1,716.2
1,265.4
Ipsos SA
France
67
9,498
1,512.8
1,383.0
Synovate
U.K.
64
5,902
884.8
776.3
270.0
SymphonylRI Group
U.S.
8
3,600
727.0
Westat Inc.
U.S.
1
1,964
455.3
—
INTAGE Inc.**
Japan
3
1,997
416.2
12.7
Arbitron Inc.
U.S.
2
1,113
395.4
5.0
The NPD Group Inc.
U.S.
14
1,135
240.1
66.4
J.D. Power and Associates*
U.S.
10
730
231.4
84.1
Video Research Ltd.**
Japan
3
393
212.0
—
IBOPE Group
Brazil
15
2,259
210.8
38.8
ICF International Inc.
U.S.
7
1,107
200.1
46.9
comScore Inc.
U.S.
18
922
175.0
32.7
Harris Interactive Inc.
U.S.
7
853
166.8
72.9
Maritz Research
U.S.
5
743
162.1
21.2
Abt SRBI Inc.
U.S.
37
819
129.4
42.6
Macromill Inc.
Japan
2
553
115.0
—
Lieberman Research Worldwide U.S.
3
368
100.8
18.5
ORC International
U.S.
6
444
98.0
34.1
Mediametrie
France
1
547
92.3
8.5
Cello Research & Consulting
U.K.
2
353
84.5
37.2
Market Strategies International U.S.
3
Total
257
80.7
13.1
107,803
$18,759.9
$10,578.6
SOURCE: Jack Honomichl, “Top 25 Global Research Organizations,” Marketing News, August 30, 2011: p. 16.
developing countries have major marketing research organizations. Of the top 25, 14 companies
grew their empires during 2007 primarily through acquisition.11 Table 2-2 lists the top 10 global
marketing research organizations.
CAREER OPPORTUNITIES IN MARKETING
RESEARCH
Marketing research offers several promising career opportunities,12 depending on one’s level of
education, experience, interests, and personality. Interesting and exciting careers are available
both within research supplier organizations—typically, external suppliers of marketing research
42
QUESTIONS
AND
PROBLEMS
services such as A.C. Nielsen, Information Resources, Inc. (IRI), J.D. Power and Associates, and
so on—and within companies that have their own research department. A brief description of
marketing research jobs, the required level of education, the real level of experience, and the
average annual compensation are provided in the Appendix of this chapter.
Demand is outstripping the research industry’s supply of high-level data analysts, also
known as marketing research methodologists. The marketing research methodologist is an individual who has a balanced and in-depth knowledge of the fields of statistics, psychometrics, marketing, and buyer behavior and applies that knowledge to describe and infer causal relationships
from marketing data. Typically, today’s marketing research methodologist has specialized in one
of the fields mentioned at the master’s and doctorate level. They will have to acquire considerable knowledge in the related fields through additional coursework, targeted in-depth readings,
and ongoing communication and interaction with peer-level specialists. However, some wellknown methodologists emerged from the engineering and hard science disciplines while others
migrated from political science, sociology, agricultural science, geology, and anthropology. A
well-rounded research methodologist would include the following coursework in his or her program of study:
■ Probability and statistics
■ Sociology
■ Multivariate statistics
■ Buyer behavior
■ Human psychology
■ Microeconomics
■ Organizational psychology
■ Marketing management
■ Psychometric measurement
■ Business communications
■ Marketing research
■ Social media marketing
The need for advanced analytical techniques has been accelerating for the last several years
as both research companies and corporate research departments become more familiar with the
power and utility of newer methodologies. A sharp dissonance between supply and demand is
fostering:
1. Erroneous execution and delivery of advanced analyses by unqualified research analysts
2. Failure to recommend advanced methodologies by independent research organizations and
consultants
Therefore, marketing researchers must work with other interested parties to remedy the situation and offer support for the marketing methodologist position. The specialty of the marketing research methodologist will assume an ever-increasing role in the business of marketing
research.13
Summary
The focus of marketing research has shifted from ad-hoc methods to collecting data and helping
managers make informed, knowledgeable decisions. The marketing decision support system is the latest
in a series of developments that help marketing managers use the information they obtain in a more
meaningful manner. Marketing research is a key component of the MDSS because it provides one of the
main inputs into the system. Marketing research can either be done in house or bought from outside
suppliers. A number of market research companies provide many services, both syndicated and
customized.
43
Questions and Problems
1.
2.
3.
a. How do marketing information systems aid
marketers in their decision making?
b. What types of information can be obtained from
marketing information systems?
What are the inherent characteristics of an effective
marketing decision support system?
A marketing manager needs to find the causes for
the decline in market share of his or her company’s
product. The manager decides to conduct marketing research.
4.
a. How should he or she go about finding a supplier of research services?
b. Suppose the manager decides to introduce the
product in Europe. A study needs to be conducted
to assess the acceptance of the product in various markets. What criteria should the manager
use in selecting a research supplier?
Identify three examples of identification of market
opportunities or marketing problems from recent
secondary data sources.
End Notes
1. Sil Seggev, “Listening Is Key to Providing Useful
Marketing Research,” Marketing News, January 22,
1982, p. 6.
2. Kenneth R. Wade, “The When/What Research
Decision Guide,” Marketing Research: A Magazine
of Management & Applications, 5, Summer 1993,
pp. 24–27.
3. “Marketing: The New Priority,” Business Week,
November 21, 1983, p. 96.
4. Cyndee Miller, “Computer Modelling Rings the
Right Bell for Avon,” Marketing News, May 9, 1988,
p. 14.
5. Paul Boughton, “Marketing Research Partnerships:
A Strategy for the ’90s,” Marketing Research: A Magazine of Management and Applications, 4, December
1992, pp. 8–13.
6. Jack Honomichl, “Top 50 Report,” Marketing News,
June 39, 2009: 12.
7. Jack Honomichl, “Top 50 Report,” Marketing News,
June 39, 2009: 12.
8. Useful sources are Greenbook: International Directory
of Marketing Research Houses and Services, New
York: American Marketing Association, annual;
Consultants and Consulting Organizations Directory,
Detroit: Gale Research Co., triennial with annual
supplements; and Bradfords Directory of Marketing
Research Agencies in the U.S. and Around the World,
Appendix
Marketing Research Jobs
Visit http://www.drvkumar.com/MR11.
44
9.
10.
11.
12.
13.
Fairfax, VA: Bradford Publishing Co., 1984. A list
of the top 300 companies that specialize in focusgroup interviewing techniques appears in the
January issues of Marketing News.
Marketing News provides a directory of software
for marketing research application.
Raymond D. Speer, “Follow These Six Steps to Get
Most Benefit from Marketing Research Consultant
Project,” Marketing News, September 18, 1981,
pp. 12–13. Marketing News publishes a directory of
international marketing research firms on a
regular basis.
Jack Honomichl, “Top firms consolidated grip on
industry,” Marketing News, August 15, 2008: H2.
Thomas C. Kinnear and Ann R. Root, Survey of
Marketing Research Chicago American Marketing
Association, 1996, Carl McDaniel and Roger Gates,
Contemporary Marketing Research, Minneapolis,
MN: West Publishing Co., 26. Marketing News on a
regular basis publishes a directory of marketing
research firms located within research users such
as Coca-Cola, AT&T, etc., in their in-house research
department.
William D. Neal, “The Marketing Research
Methodologist,” Marketing Research, Spring 1998,
Vol. 10, 1, pp. 21–25.
THE MARKETING
RESEARCH PROCESS
3
Learning Objectives
■ Be familiar with the various stages of the marketing research process.
■ Highlight the importance of the problem/opportunity identification stage of the research
process.
■ Understand the issues related to hypotheses development.
■ Explain the concept of value of information, and its role in deciding when marketing research
is beneficial.
■ Introduce the international marketing research process.
How is the market research project conceived, planned, and executed? The answer, in part, is
through a research process, consisting of stages or steps that guide the project from its conception through the final analysis, recommendation, and ultimate action. The research process provides a systematic, planned approach to the research project and ensures that all aspects of the
research project are consistent with each other. It is especially important that the research design
and implementation be consistent with the research purpose and objectives. Otherwise, the
results will not help the client.
The research process is described in this chapter and Chapter 4. This chapter provides an
overview of the research process, a discussion of the research purpose and research objectives,
and a consideration of the value of research information. Negative findings are as valuable as
positive ones. In fact, they are often more revealing, as they provide valuable insight into customers’ psyches. Today, the research process has evolved to encompass decision making. This
combined process transforms mundane marketing research to marketing intelligence. Chapter 4
gives an overview of the research design and its implementation. Together, these two chapters
are the foundation for the rest of the book.
OVERVIEW OF THE MARKETING RESEARCH
PROCESS
Research studies evolve through a series of steps, each representing the answer to a key question.
1. Why should we do research? This establishes the research purpose as seen by the management
team that will be using the results. This step requires understanding the decisions to be made
and the problems or opportunities to be diagnosed.
2. What research should be done? Here the management purpose is translated into objectives
that tell the managers exactly what questions need to be answered by the research study
or project.
3. Is it worth doing the research? The decision has to be made here about whether the value
of the information that will likely be obtained is going to be greater than the cost of collecting it.
45
CHAPTER 3 THE MARKETING RESEARCH PROCESS
4. How should the research be designed to achieve the research objectives? Design issues include the
choice of research approach—reliance on secondary data versus conducting a survey or
experiment—and the specifics of how to collect the data. Chapter 4 deals with how to
approach these issues.
5. What will we do with the research? Once the data have been collected, how will it be analyzed,
interpreted, and used to make recommendations for action?
The necessary steps are linked in a sequential process (see Figure 3-1). Although the steps
usually occur in this general order, we must emphasize that “early” decisions are always made
by looking ahead to “later” decisions. The early decisions are constantly being modified to
MARKETING PLANNING AND INFORMATION SYSTEM
Planning system
Strategic plans
• Tactical plans
Information system
• Databases
• DSS
•
1. AGREE ON RESEARCH PROCESS
•
•
•
Problems or opportunities
Decision alternatives
Research users
2. ESTABLISH RESEARCH OBJECTIVES
•
•
•
Research questions
Hypotheses
Boundaries of study
3. ESTIMATE
THE VALUE OF
INFORMATION
No
Is benefit > cost?
Yes
4. DESIGN THE RESEARCH
•
•
•
•
Choose among alternative research approaches
Specify the sampling plan
Design the experiment
Design the questionnaire
5. COLLECT THE DATA
6. PREPARE AND ANALYZE THE DATA
7. REPORT THE RESEARCH RESULTS AND
PROVIDE STRATEGIC RECOMMENDATIONS
46
FIGURE 3-1
The marketing research process.
DO NOT CONDUCT
MARKETING RESEARCH
THE PRELIMINARY STAGES
OF THE
MARKETING RESEARCH PROCESS
account for new insights and possibilities presented by later decisions. Also, the steps do not
function in isolation. Rather, they are embedded in the ongoing planning process of the business, which culminates in the development of strategies, programs, and action. This planning
process provides the purposes of the research. In turn, planning is supported by the information
system, which (1) anticipates the type of information required by decision makers and (2) organizes data that have been collected to ensure their availability when needed.
The development of a research purpose that links the research to decision making, and the
formulation of research objectives that serve to guide the research, are unquestionably the most
important steps in the research process. If they are correct, the research stands a good chance of
being both useful and appropriate. If they are bypassed or wrong, the research almost surely will
be wasteful and irrelevant. These aspects of research, too often neglected by managers, will be
discussed in detail in this chapter. The next chapter deals with research design; the chapters in
Part II discuss the various methods to collect data; and the chapters in Part III of the book deal
with analysis and interpretation of the data.
THE PRELIMINARY STAGES OF THE
MARKETING RESEARCH PROCESS
Step 1—Research Purpose
Research problems are more likely to be poorly defined, only partially understood, and missing
possible decision alternatives that should be analyzed. Defining problems accurately is a combination of data and judgment that demands real thought and effort. Problems, opportunities,
and “nonproblem” situations are closely related to structure. Together they make up a family of
gaps. The concept of analyzing the gaps as problems is based on:
1. Recognizing/understanding a problem. A problem is a gap between what was supposed to happen and what did happen between our objective and our accomplishment. Three elements
are required to recognize a problem:
■ Something must be expected to happen.
■ Feedback must be received on what actually happens.
■ Expectations and feedback must be compared.
2. Knowing where and when the gap or problem occurred. Once a problem is defined, it is easier to approach the cause and solution to the gap(s), in accordance with the level of detail
of the analysis. In the end, problem definition is and will always be a creative act, a balance between thorough research and intuition. Problem definition is best thought of as a
solution definition—the selection of a domain is likely to be rich in ideas to solve the
problem. Problem definition is a creative act. The payoff from good marketing definition
is enormous—nothing else we do has so much leverage on profit.1
Seldom will research problems come neatly packaged with obvious information requirements, clear-cut boundaries, and pure motives on the part of the decision makers. Launching a
research study with such shaky inputs is a recipe for producing unusable findings and unhappy
clients. It is in the best interest of both the researcher and the managers paying for the research
to be sure that the research purpose is fully understood. One of the hallmarks of a competent
researcher is the ability to get to the heart of the management problem.
Consider the seemingly straightforward request by the chairperson of an association of community merchants for a research project. The objective of this project was to help reduce the
propensity of residents in the community to do their shopping in two nearby communities.
Clearly, the purpose of the research was to identify and evaluate various ways to increase the
local merchants’ share of shopping by residents.
Further probing, however, revealed that the statement of the problem was at least partially
inaccurate. Only late in the research process was it learned by the researcher that the chairperson
47
CHAPTER 3 THE MARKETING RESEARCH PROCESS
was having real difficulty convincing the other local merchants that there was a serious enough
outflow of local trade to warrant joint action to reverse the flow. This certainly changed the purpose of the research. Now the researcher would have to measure the level of retail trade outflow,
in addition to finding the reasons for the outflow. This required a major change in the research
design, but had the change not been made the results would have been of little value to the client.
The research purpose comprises a shared understanding between the manager and the
researcher of:
1. Problems or opportunities to be studied
■ Which problems or opportunities are anticipated?
■ What is the scope of the problems and the possible reasons?
2. Decision alternatives to be evaluated
■ What are the alternatives being studied?
■ What are the criteria for choosing among the alternatives?
■ What is the timing or importance of the decision?
3. Users of the research results
■ Who are the decision makers?
■ Are there any covert purposes?
Problem or Opportunity Analysis. In analyzing problems or studying opportunities, identifying loyal customers and doing targeted marketing are very important. Marketing Research in
Action 3-1 reiterates the importance of market research and databased marketing in creating “top
of mind awareness” of their products.
Research often is motivated by a problem or opportunity. The fact that sales are below
expectations in the East might be a problem requiring research. The fact that people are consuming fewer sweets might be a problem or a potential opportunity for a candy company.
Increased leisure time might be viewed as an opportunity by a recreation-oriented organization.
In such cases the research purpose should specify the problem or opportunity to be explored. Identifying and defining the problem or opportunity is a crucial first step in the marketing research
process. Especially in situation analysis contexts, exploratory research is needed to identify
problems and opportunities. What sales areas are showing weak performance? What segments
represent opportunities because they are dissatisfied with current products or because they are
underusing the product? Even in exploratory research, however, it will be helpful to identify
MARKETING RESEARCH IN ACTION 3-1
What Do Today’s Buyers Want?
Today’s buyers have created changes for all businesses. They are
■ Skeptical—their trust has been broken by every industry.
■ Cautious—they take a lot of time to weigh decisions.
■ Tired of selling and sales pressure—they are weary of spam and telephone sales.
■ Busy—they want everything quick and easy.
■ Confused—they are bewildered with the variety of offers.
Today’s buyers have access to information because of the Internet, and they can make better decisions
because of the global market available for choice. Marketers have to create a “top of mind awareness”
of their products. To do this, they have to constantly inform buyers about their products without irritating them. This means they have to identify their most loyal customers and do a great deal of targeted
marketing. This is possible only through marketing research and databased marketing.
Once marketers have identified their customers, it’s very easy to find buyers with a similar profile
and generate loyal, profit-generating customers.
48
SOURCE: findmorebuyers.com/page.cfm/11, January 2003.
THE PRELIMINARY STAGES
OF THE
MARKETING RESEARCH PROCESS
MARKETING RESEARCH IN ACTION 3-2
Best Buy Goes Lean
Best Buy, with intense competition from competitors, has slashed capital spending for 2009 budget by
half along with buy outs and layoffs. Given Best Buy’s emphasis on its personnel, the suggestion of layoffs demonstrates a change in long-held priorities.
Best Buy faces intense competition mainly from retailers such as Wal-Mart, Target, and Costco coming after mainstay businesses including television, digital cameras and gaming. For instance, television
shoppers at Best Buy are going for less expensive sets, especially for their second or third TVs. Further,
consumers replacing their TVs are settling for 34-inch screens from 42-inch screens. Best Buy’s competitors use fewer employees to run the electronics business throughout and continue to figure out ways
to sell electronics right off the rack rather than through a sales clerk.
By becoming leaner, Best Buy is preparing itself for a competitive reality where consumers are
becoming more considered about spending and where discounters and warehouse clubs are the major
competition.
SOURCE: Mike Duff, “Best Buy Goes Leaner As Customers Waver, Competitors Press,” BNET Retail, December 19, 2008.
the nature of the problem or opportunity that is motivating the research. Further, the goal
should be to move from exploratory research to research more focused on a decision. Marketing
Research in Action 3-2 illustrates the process of changing marketing approach by reacting to
market situations.
The manager needs to make certain that the real problem is being addressed. Sometimes the
recognized problem is only a symptom, or perhaps merely a part of a larger problem. A sobering
illustration of this is the plight of Compton Corp.,2 a manufacturer of capital equipment costing
between $10,000 and $25,000. The company was dominant in its market, with a share as large as
the next two biggest competitors. All the companies sold their equipment through a network of
independent distributors, each of which sold the products of at least two competitors. For several
years this market leader had been losing share. In an attempt to reverse the trend, the company
changed advertising agencies. When the new agents funded a study of end users, they found to
their surprise that the previous agency had done a superb job of creating awareness and favorable
attitudes. However, many of the equipment purchasers who favored Compton were actually buying the competing brands. This problem had little to do with the performance of the advertising
agency. A new study, oriented toward the distributors, found that Compton’s distributor-relations
program was very weak relative to its competitors. One competitor emphasized sales contests,
another offered cash bonuses to salespeople, and a third was particularly effective with technical
sales assistance directed to difficult accounts. Not surprisingly, these factors influenced the distributors when they were asked for advice, or when the prospective purchaser did not have a firm
commitment to Compton equipment.
In this case, the real problem ultimately was isolated, but only after much time and energy
had been directed toward the wrong problem. When defining the problem, it is important to
think broadly about the possible causes or influential variables. This may justify a separate
exploratory research study. Further, what appears to be a genuine problem or opportunity may
not be researchable. For example, if a company that manufactures washing machines is interested in determining the replacement rates for all machines sold within the last three years, it
may not be worthwhile to pursue the issue. Since most household washing machines have a life
span ranging from five to 10 years, the problem of identifying the replacement rate for working
machines sold within the last three years may be a nonresearchable problem.
Decision Alternatives. For research to be effective, it must be associated with a decision. Marketing research is committed to the principle of utility. In general, if research is not going to have
an effect on decisions, it is an exercise in futility. The researcher should always be sensitive to
the possibility that either there are no decision alternatives—and therefore no decision—or that
the research findings will not affect the decision, usually because of resource or organizational
49
CHAPTER 3 THE MARKETING RESEARCH PROCESS
constraints. In such circumstances, the research will have no practical value and probably should
not be conducted.
When a decision potential does exist, it is important to identify it explicitly, because the research
then can be designed for maximum effectiveness. For example, researchers frequently are asked to
assess the potential of a market that is not familiar to the company. But what are the decisions the
manager faces? Is the manager thinking of acquiring a company serving that market? Has the lab
produced a new product that might be sold as a component to the industry serving that market?
The answers will have a significant influence on the design of the research.
A most useful way to clarify the decision motivating the research is to ask: (1) What alternative actions are being considered? (2) What actions would be taken, given the various feasible
outcomes of the research? This line of questioning can be very enlightening for the decision
maker, as well as for the researcher, in terms of clarifying exactly what the research can accomplish. The story in Marketing Research in Action 3-3 illustrates how both can learn from a focus
on decisions.
Sometimes the decision involved is highly specific. A copy test is used to select a copy alternative. A concept test is employed to determine if a concept should be developed further. Sometimes the decision can be very general. What markets should be the primary targets of our organization? Should our marketing program be changed? It is desirable to be as specific as possible,
because the research purpose then will be more effective in guiding the development of the
research design. However, even if the decision is necessarily general, it needs to be stated clearly.
Criteria for Choosing Among Alternatives. It is essential for the researcher to know how the
decision maker will choose among the available alternatives. Suppose a product manager is considering three possible package redesigns for a health-care product with declining sales. This
would seem to be a straightforward research undertaking, as the decision alternatives are completely specified. However, the product manager could use some or all of the following criteria
to choose the best of the three alternative packages:
1.
2.
3.
4.
5.
Long-run sales
Trial purchases by users of competing brands
Amount of shelf space assigned to the brand
Differentiation from competitive packages
Brand-name recognition
MARKETING RESEARCH IN ACTION 3-3
50
Political Campaign Research
The meeting between Hugh Godfrey and two project directors from Pollsters Anonymous, a well-known
survey research company, had taken a surprising turn. Here were two researchers suggesting that no
research be undertaken.
Godfrey was campaign manager for John Crombie, a university professor and erstwhile Democratic
challenger of the Republican incumbent for the local House of Representatives seat. He and his candidate were anxious to undertake a program of research. They thought it would be a good idea to take
surveys in May and September (five months and six weeks prior to the election) of voter awareness of
the candidate, attitudes toward him, issue salience, and intentions to vote. The results would be helpful in clarifying the candidate’s position and deciding on media expenditures. Positive results would
be useful in soliciting campaign contributions, which loomed as a big problem.
During the meeting the researchers had asked what Godfrey expected to find. He was sure that the
initial survey would reveal low awareness, and would confirm other information he had that there was
a low level of voter registration among Democrats in the area. The next question was whether any foreseeable results would persuade him not to spend all his available resources on a voter registration drive.
He also admitted that the preliminary estimate of $6,000 for a May survey was a large chunk of his
available funds. In fact, he was thinking, “With the money I would spend on the survey, I could hire
enough canvassers to get at least 1,500 to 2,000 registrations.”
THE PRELIMINARY STAGES
OF THE
MARKETING RESEARCH PROCESS
The researcher and decision maker need to discuss all possible criteria in advance, and
choose those that are appropriate. If the criterion for comparison is long-run sales results, the
research approach will be much more elaborate than if the choice is based simply on brand-name
recognition.
Timing and Importance. Timing and importance are always pivotal issues in the research
process. How crucial is the decision? If the wrong decision is made, what will be the consequences? Obviously, the decision to “go national” with a new government program represents
a much larger commitment than the decision to pursue a new program idea a bit further. Other
questions concern the timing of the decision. What is the time pressure on the decision? Is information needed quickly, or is there time to develop an optimal research design?
Research Users—Decision Makers. When the research results will be used to guide internal problem solving, the researcher must know the objectives and expectations of the actual decision makers. The bigger the problem, the more difficult this becomes, for not only are a large number of
people likely to be involved, but the contact person may simply be acting as a liaison whose interpretation of the problem and the need for research may be secondhand. The major benefit from
making an effort to reach all the decision makers is that the research purpose is likely to be specified more adequately. These contacts also will tell the researcher (especially an outside supplier who
is called in to undertake the work) a good deal about the resources that are available to deal with
the problem. This is very helpful in developing a realistic proposal.
Increasingly, marketing research is entering the public domain, which introduces a new set of
users who frequently have very different criteria for evaluating research results. For example:
■ A public utility presents a research study to a regulatory body in support of a request for a
rate change or the introduction of a change in service level.
■ An industry trade association conducts research designed to influence proposed legislation or
trade regulations. The Direct Mail Marketing Association has sponsored a study of mail-order
buyers in response to a proposed Federal Trade Commission order that would require sellers
to offer a refund if they could not ship the ordered goods within a month.
■ A regional transit agency wants to build public support for the continuation of an experi-
mental program involving “dedicated” bus lanes (part of a road or highway on which no automobile traffic is permitted). The research demonstrating the effectiveness of the program is to
be presented to various public bodies and citizen groups.
In most cases, the research in the above examples will be used to support a decision alternative.
However, examination of the results often is conducted in an adversarial setting, which means
more criticism of shortcomings and necessitates a higher quality of research.
Overt and Covert Purposes. It would be naive to presume that research is always conducted to
facilitate rational problem-solving activity or that the decision maker always will be willing or able
to share reasons for initiating the research. As discussed in Chapter 1, there are times when the
main purpose of marketing research is to serve someone’s organizational goals or for other unethical purposes. None of these abuses can be condoned. Often they are specifically prohibited by
industry codes of ethics. When they are not, one’s moral standards become the compass for deciding what is right.
Step 2—Research Objective
The research objective is a statement, in as precise terminology as possible, of what information
is needed. The research objective should be framed so that obtaining the information will ensure
that the research purpose is satisfied.
51
CHAPTER 3 THE MARKETING RESEARCH PROCESS
Research objectives have three components. The first is the research question. It specifies the information the decision maker needs. The second and third elements help the researcher make the
research question as specific and precise as possible. The second element is the development of
hypotheses that are basically alternative answers to the research question. The research determines
which of these alternative answers is correct. It is not always possible to develop hypotheses, but
the effort should be made. The third is the scope or boundaries of the research. For example, is the
interest in current customers only or in all potential customers?
Research Question. The research question asks what specific information is required to achieve
the research purpose. If the research question is answered by the research, then the information
should aid the decision maker.
An illustration comes from a company in the toiletries and cosmetics business, which was
interested in acquiring a smaller firm with an apparently complementary product line. One anticipated benefit of the acquisition was the opportunity to eliminate one of the sales forces. The
purpose of the research was to assess whether the company could use its existing sales force to
distribute the products of the acquired company. The corresponding research objective was to
determine how much the retail distribution patterns of the two companies overlapped. There
was some preliminary evidence suggesting (that is, hypothesizing) that distribution coverage
would differ by geographic area and store type. The resulting study found that there was very
little overlap, because the acquiring company emphasized major metropolitan areas, whereas the
other company was represented largely in smaller cities and suburbs.
It is possible to have several research questions for a given research purpose. Thus, if the
purpose is to determine if a specific advertisement should be run, the following research questions could be posed:
■ Will the advertisement be noticed?
■ Will it be interpreted accurately?
■ Will it influence attitudes?
These questions correspond to the criteria used to evaluate the advertising alternatives. Similarly,
if the purpose is to determine how to improve the services of a bank, possible research questions might be
■ What aspects of the current service are customers most pleased with, and with which are they
most dissatisfied?
■ What types of customers use the various services?
■ What benefits do people seek from banks?
52
Each of these questions should pass the test of being relevant to the purpose. For example, if
customer types are identified that use a service such as traveler’s checks, it may be possible to
modify that service to make it more convenient or attractive to them.
Sometimes the researcher can select a major objective and some supporting objectives. An example is a study conducted for the U.S. Department of Defense to address the problem of declining
strength of the National Guard and the reserve components such as the Army Reserve.3 The research
purpose was to determine what job characteristics (product dimensions) would increase the enlistment levels and the reenlistment levels of various demographic types. Job characteristics such as salary,
fringe benefits, educational opportunity, travel, job image, and hair regulations were among the possible policy variables that could be adjusted. The overall objective of the study, to examine motivation
factors in enlistment and reenlistment, led to several supporting objectives. The first of these was to
measure young people’s propensity to serve (or reenlist). The second was to determine current perceptions of the reserve in terms of 12 key attributes. The third was to determine the relative importance of the 12 key job attributes that could provide the basis for influencing young men and women
to join and remain in the service.
THE PRELIMINARY STAGES
OF THE
MARKETING RESEARCH PROCESS
The researcher will always try to make the research question as specific as possible. Suppose the research question as to which customer types use the various bank services could be
replaced by the following research question: What are the lifestyle and attitude profiles of the
users of credit cards, automatic overdraft protection, and traveler’s checks? This increase in
specificity would aid the researcher in developing the research design by suggesting whom to
survey and what questions to include. The role of the research objective is to provide guidance
to the research design. The more specific the research question is, the more practical guidance
will be provided.
When a research question is set forth, it is sometimes difficult to realize that the question
can and should be made more specific. The remaining two elements of the research objective—
hypothesis development and the research boundaries—provide exercises to help the researcher
make the research question more specific.
Hypothesis Development. A hypothesis is a possible answer to a research question. The
researcher should always take the time and effort to speculate as to possible research question
answers that will emerge from the research. In doing so, the fact that everyone already knows
the answer sometimes becomes apparent. More often, the effort will add a considerable degree
of specificity to the research question.
A hypothesis could speculate that sales are down in the Northeast because the level of competition has been abnormally high there during the past two months. Such a hypothesis provides
considerable detail to a research question that asks what the problem is in the Northeast. It guides
the research by ensuring that competitive promotions are included in the research design. One
important role of a hypothesis is to suggest variables to be included in the research design—in
this case, competitive promotion.
A research problem might be to estimate the demand for a new product. The hypothesis that
the product will do well in the North but not in the South adds the concept of geographic location to the problem. It suggests that the sampling plan should include people from both regions.
If the hypothesis suggests that the product will not do well in the South because it is not compatible with the Southern lifestyle, it becomes evident that the research should measure not only
purchase intentions, but also how the product would be used.
Normally, there will be several competing hypotheses, either specified or implied. If all the
hypotheses were known in advance to be true, there would be little reason to conduct the
research. Thus, one objective of research is to choose among the possible hypotheses. A good
illustration of the role of competing hypotheses is the problem faced recently by a satellite television company. A satellite TV company picks up TV and radio signals and “pipes” the highquality signals directly into subscribers’ homes via satellite. This company provided service to
75 percent of the households within its total service area. The problem facing the company was
that there were several areas where the penetration rate was far below average. The population
in these areas represented about 15 percent of the total service area. Bringing these areas closer
to the average would improve profitability significantly. Before remedial action could be taken,
however, it was necessary to establish the reasons for the low penetration. Various reasons were
suggested by management, including the following:
1.
2.
3.
4.
5.
6.
Good television reception is available.
There is a very large transient population.
Residents have had poor previous experience with satellite service.
The price is too high, given the incomes in the area.
The sales force coverage has been inadequate.
A large percentage of the residents of the area are in age or social class groups that watch
little television.
The challenge for the researcher is to devise a research approach that will gather information that can test each of these hypotheses. Hypotheses are not appropriate for all situations. As
the upcoming discussions on exploratory research in Chapter 4 will make clear, there may be
53
CHAPTER 3 THE MARKETING RESEARCH PROCESS
insufficient information for developing hypotheses. There are also times when the most reasonable hypothesis statement is simply a trivial restatement of the research question. For example:
Research
Question: Will the advertisement attract attention?
Hypothesis: It will attract attention.
In such cases the hypothesis will not add anything to the research and should simply be
omitted. Hypothesis development should not be viewed as an item on a checklist to be quickly
satisfied, but rather as an opportunity to communicate information and to make the research
question more specific.
How does the researcher generate hypotheses? The answer is that whatever information is
available is used to speculate on which answers to the research questions are possible and which
are likely. The researcher can use three main sources of information to develop hypotheses, as
Figure 3-2 illustrates. First, the researcher can draw on previous research efforts; in fact, it is not
uncommon to conduct exploratory research to generate hypotheses for future large-scale research
efforts. The research purpose might be deciding whether to conduct the large-scale studies.
A second source of hypotheses is theory from such disciplines as psychology, sociology, marketing, or economics. Thus, economic theory might suggest the importance of price in explaining a loss of retail sales. Marketing theory could indicate that distribution is important in predicting new-product acceptance. The use of attitude as a measure of advertising impact might
be suggested by psychological theory.
A third and perhaps the most important source of hypotheses is the manager’s experience
with related problems, coupled with a knowledge of the problem situation and the use of judgment. This source is illustrated by the manufacturer who has discovered an unusual increase in
selling costs. Past experience with similar problems, plus a preliminary investigation into the reasons for the problem, point to an increase in the proportion and number of small orders received.
The tentative hypothesis is: Small orders (suitably defined) have increased in both number and
proportion, and this increase, coupled with a higher cost of processing these orders, has raised
selling costs. The research could then be directed at the questions of (1) the extent of increase in
small orders (and the reasons for the increase) and (2) the additional unit costs involved in processing orders of different sizes.
Research Boundaries. Hypothesis development helps make the research question more precise.
Another approach is to indicate the scope of the research or the research boundaries. Is the interest in the total population restricted to men, or to those on the West Coast? Is the research question
Source
•
•
•
Research
purpose
Theory
Management experience
Exploratory research
Research
question
Hypothesis
Research
objective
54
FIGURE 3-2
Hypothesis development.
Research
design
THE PRELIMINARY STAGES
OF THE
MARKETING RESEARCH PROCESS
concerned with the overall attitude toward the proposed new automobile, or is it necessary to learn
customer attitudes about trunk space, handling, gas economy, styling, and interior appearance?
Much of the dialogue between the researcher and the decision maker will be about clarifying the boundaries of the study. For example, a manager may wish to study the effects of the
European Union’s core industrial policy inspired by the Treaty on European Union, which came
into force in 1993.4 During the process of hypothesis development, the possible effects may be
isolated. This still leaves a number of areas of ambiguity. What is meant by “condition”—
profitability, competitive position in world markets, labor relations? How is the “industry” to be
defined? What geographic areas are to be considered? What time period is to be appraised?
A final question of research scope regards the desired precision or accuracy of the results.
This will, of course, depend on the research purpose. If a multimillion-dollar plant is to be constructed on the basis of the research results, a high degree of accuracy will be required. If, however, the decision involves the investment of a small sum in research and development on a new
product idea, then a crude judgment as to the potential of the product would be acceptable.
Step 3—Estimating the Value of Information
Before a research approach can be selected, it is necessary to have an estimate of the value of
the information—that is, the value of obtaining answers to the research questions. Such an estimate will help determine how much, if anything, should be spent on the research.
The value will depend on the importance of the decision as noted in the research purpose,
the uncertainty that surrounds it, and the influence of the research information on the decision.
If the decision is highly significant in terms of the investment required or in terms of its effect
on the long-run success of the organization, then information may have a high value. However,
uncertainty that is meaningful to the decision also must exist if the information is to have value.
If the outcomes are already known with certainty, or if the decision will not be affected by the
research information, the information will have no value.
To illustrate and expand on these concepts, consider the simplified examples in Figure 3-3. In
case A, the decision to introduce a new product is shown as a decision tree. The first two branches
represent the decision alternatives—to introduce the product or to decide not to introduce it. The
second branch represents the uncertainty. Our descriptive model indicates that if the product is
successful, a profit of $4 million will result. The indication is that there is a probability of 0.6
(obtained from prior experience) that the product will be successful. However, if the product is
not successful, the profit will be only $1 million, an event that will occur with probability 0.4.
These subjective probabilities have been calculated based on prior knowledge of the situation.
How much should we be willing to pay for perfect information in this case? If someone could
tell us, in advance and with certainty, whether the product would be successful, how much would
0.6
Success
0.4
Failure
0.6
Success
0.4
Failure
$4 million
Introduce
Case A
Do not
introduce
$1 million
$0
$4 million
Introduce
Case B
Do not
introduce
–$2.5 million
$0
FIGURE 3-3
Illustrative decision models.
55
CHAPTER 3 THE MARKETING RESEARCH PROCESS
we pay for that information? The correct answer is nothing! The fact is that our decision would
be the same regardless of the information. We would introduce the product, for even if the product were not well accepted, we would still make the $1 million. In this case, not only is the decision insignificant to the organization, it is nonexistent. There is only one viable alternative. Without alternatives there is no decision contest, even if uncertainty exists; therefore, there is no need
for additional information.
In case B, the estimate is that, if the product is not successful, a loss of $2.5 million will occur.
Since the expectation of the new product’s eventual performance is still, on balance, positive, the
product would be introduced.* In this case, however, perfect information now would have value.
If we knew in advance that the product would not be accepted, we would decide against introducing it and save $2.5 million. Since our best estimate of the probability of the product not
being accepted is .4, the value of the information would be .4 times $2.5 million, or $1 million.
Thus, if this decision contest could be repeated many times, perfect information would save us
$2.5 million about 40 percent of the time and would save us nothing (since it would not alter
our decision) about 60 percent of the time. On average, it would save us $1 million. By spending money on research, we might improve our knowledge of how the product will be accepted.
But market research is unlikely to be as good as perfect information, and therefore its value will
be less than $1 million. Obviously, if the cost associated with an unsuccessful product were lower,
or if the probability of an unsuccessful product were smaller, the value of information would be
less. (The Appendix to this chapter extends this example to include the possibility of using a
concept test to predict whether or not the product will succeed. A method is developed to determine the value of the concept test to the decision maker.)
PLANNING A NEW HMO
To see how a research purpose and a set of research objectives are developed, let us join a meeting that took place at the Fraser General Hospital in September 2002.
The five doctors had a dilemma. They had spent a useful morning confirming that their hospital had the resources to operate a health maintenance organization (HMO). These resources were
substantial, as would be expected in a big teaching and research hospital with a strong regional
reputation. The concept of an HMO is explained briefly in Marketing Research in Action 3-4.
Leading HMOs are experimenting with the demi-science called “alternative medicine,” hoping to cure ills when regular medicine fails. This is a startling trend. Many doctors still view
some alternative methods with great skepticism—the word “quackery” is bandied about—and
health maintenance organizations have a reputation for shunning all except the most reliable
treatments. While most HMOs already offer chiropractic coverage, a number are adding
acupuncturists and massage therapists, along with practitioners called naturopaths who use
MARKETING RESEARCH IN ACTION 3-4
What Is an HMO?
There are two basic kinds of health coverage. The health maintenance organization is the best example of
the prepaid group-practice type. This involves fixed monthly payments directly to a group of doctors or
a clinic that is then responsible for all the health needs covered by the plan. The other type of coverage is
the insurance type, which involves a company that collects premiums from subscribers, who can go to any
clinic or doctor they choose. The insurance company pays for the services covered by the policy. In both
cases there may be limits to the coverage of such items as hospitalization, drugs, and office visits. The big
drawback of the HMO is restriction of the choice of physicians and hospitals to those affiliated with the
HMO. Generally, the total annual cost of an HMO to the consumer is lower than for group insurance plans.
One reason is that the flat-fee formula discourages doctors from hospitalizing patients for longer than necessary. Also, the emphasis on preventative care produces fewer seriously ill patients.
56
*The expected value of introducing the product would be 0.6(4M) ⫹ 0.4(⫺2.5M) ⫽ 1.4M.
PLANNING
A
NEW HMO
herbal remedies, relaxation therapies, yoga, and more. The health plans are offering to cover
this care just as they do treatment by cardiologists and pediatricians.
The September 10, 2002, meeting was one in a long series of informal talks, investigations,
and efforts to build support for the idea within the hospital. These efforts had brought the doctors to a critical point: Their problem was how to determine whether there was enough market
demand in the region to support another HMO. Although each member of the planning group
was convinced of the prospective benefits to the community, the hospital teaching program, and
the bed utilization rate, they knew they needed persuasive evidence before the hospital trustees
would provide start-up funds. After all, the trustees were even reluctant to approve funds for
initial planning. What would it take before they would approve an initial investment in excess
of $400,000?
The trustees were not the only hurdle. Each of the doctors in the HMO planning group knew
at least one colleague who was openly skeptical because of the presence of competitive healthcare programs. There was a well established HMO about 8 miles away, plus several clinics operating on a fee-for-service basis. Many of the clinic doctors had privileges at the Fraser General
Hospital.
At lunch, the planning group was joined by Herb Ellis, a partner in a local research and consulting firm. He had been invited by one of the doctors who knew him socially. During lunch,
the doctors enthusiastically described the HMO concept, what it meant to subscribers—especially
those on low incomes—and some of the innovations, such as consumer inputs to operations and
procedures.
After lunch, John Akitt, a surgeon and the originator of the HMO proposal, reviewed some
of the tentative decisions that would influence the market analysis. First, they intended to offer
fairly comprehensive services from the beginning, including internal medicine, pediatrics, surgery, orthopedics, obstetrics, and gynecology. However, they were not sure whether they should
also offer dentistry and optometry at an additional fee. Such an array of services was felt to be
competitively necessary, but when combined with a high doctor–enrolled patient ratio to ensure
a high service level, the result was very high fixed costs.
All the doctors agreed that one way to keep initial costs down was to concentrate on larger
employer groups. This meant that a two-stage marketing effort would be required—first to get
the employer to agree to offer the HMO plan as a subsidized benefit and then to persuade the
employees to switch from their present health-care plan. The largest employer in the vicinity was
the state university, 2 miles away, with over 13,000 faculty, staff, and married students who could
join a health plan. The university’s vice president of finance appeared committed to offering the
HMO plan. There were three other large employers nearby, representing an additional 20,000
possible enrollees. However, they had been much less enthusiastic about offering the Fraser Hospital HMO as a benefit, and implied that they would not consider it until it had been operating
successfully for a few months. This was upsetting to the doctors, for a sizable enrollment base
had to be generated quickly if the HMO was to achieve its objectives. Still, if there was a significant adoption of the HMO within the university, they felt they could get close to the
breakeven target in the short run.
As the discussion progressed, Ellis realized that the extent and rate of acceptance of the HMO
would depend on the marketing effort and the fee level. Marketing was a real problem, since
solicitation of patients was, strictly speaking, unethical. This probably did not prevent personal
selling and other communications efforts to explain the differences between fee-for-service and
HMO plans. Pricing seemed to be a big factor; the prepaid feature of the HMO meant that
prospective enrolled would see a bigger monthly salary bite, although over the long run total
health costs would be lower. Since a large proportion of costs were fixed, a high enrollment target could help to lower fees. Otherwise, to lower fees it would be necessary to reduce services,
which would make the HMO less attractive. Without knowledge of the price sensitivity of the
market, it would be hard to establish a fee structure. But who was the target market? On this
the doctors had many conflicting opinions and the meeting became quite heated. Was the HMO
concept most attractive to young families, older families with or without children, or enrollees
in the competitive HMO? No one was sure just how big an area an HMO could serve. Although
57
CHAPTER 3 THE MARKETING RESEARCH PROCESS
some existing HMOs attracted only people within a 10-mile radius of the hospital, there was a
feeling that convenience and driving time were more important than distance.
At this point Herb Ellis felt that he had some understanding of the doctors’ marketing problems. Rather than discuss research approaches, he asked for another meeting at which specific
research purposes and objectives could be discussed. These would be the basis for a research
proposal. He already knew that a big constraint on his planning would be the available budget.
There was no way that more than $10,000 could be found to finance marketing research, no matter what the economic value of the research information to the hospital.
Prior to his second meeting with the planning committee, Herb Ellis did some background
reading on the HMO concept, talked to several doctors in private practice, and informally interviewed four members of the competitive HMO. With this background he was ready to discuss preliminary research purposes and objectives, which could be used to guide the design of the study.
During the second meeting with the hospital planning committee, quick agreement was
reached that the primary purpose of the study was to address the decision as to whether the
proposal for an HMO should be pursued to the point of making major investments in its implementation. The following research objective consists of the research question and a statement of
the study scope.
Research
Question: What is the demand for the new HMO?
Scope: Limited to students, staff, and faculty of the university.
The study was limited to university students, staff, and faculty for several reasons. First, the
university administration was favorably disposed toward the plan, giving it the best chance of success in that environment. If support from that group was not in evidence, then the prospects would
be dim in other organizations. Second, the budget limitation made it unlikely that any worthwhile
research could be conducted with more than one organization. No geographic limits were placed
on the study as it was thought that distance from the home to the HMO would have only a weak
influence on individual interest in the proposed HMO.
Much more time was spent on developing the supporting purposes and objectives, with Ellis
constantly challenging the usefulness of each proposed purpose and objective. Because of the
tight budget constraint, he was fearful that overly ambitious objectives would be difficult to
achieve with the alternative research designs he had in mind. Finally, the following set of supporting purposes and objectives was developed.
Research
Purpose: What target market segments should the HMO emphasize?
Objective: Identify the market segments most interested in the proposed HMO. Estimate their
probable rate of utilization of medical services from their past medical experience.
Purpose: What services should be provided at what price level?
Objective: Identify the attributes or characteristics of health plans that would have the greatest
influence on an individual’s choice among alternatives.
58
During the meeting a number of hypotheses were advanced as to who was most likely to be
interested in the plan. Of course, they would have to express strong interest in the plan as described
to them. In addition, good prospects would be those who were dissatisfied with the coverage or quality of their present plan, did not have a long-standing relationship with a family doctor, had favorable attitudes toward the Fraser General Hospital, and were not enrolled in other plans through their
spouses. At the end of the meeting, the chairperson indicated a need for a proposal in time for the
trustees’ meeting on October 3. If the proposal was approved, they would need the results of the
study no later than the first week in February.
THE INTERNATIONAL MARKETING RESEARCH PROCESS
THE INTERNATIONAL MARKETING RESEARCH
PROCESS
As we mentioned in Chapter 1, the basic functions of marketing research and the various stages
in the research process do not differ between domestic and international research. The international
marketing research (IMR) process, however, is much more complicated than the domestic research
process. IMR is more complicated because of the necessity to ensure construct, measurement,
sampling, and analysis equivalence before any cross-cultural study is conducted. A thorough
research of the proposed international market is very important before launching a new product
or service. Although it is complex, it can be an extremely beneficial process. To avoid high-profile
mistakes in international marketing research, there are some considerations to be made:
1.
2.
3.
4.
Profile your target customers and clients.
Interview target segments to assess how well they match your preconceived ideas.
Hire local researchers who know the costs and methods that are workable in local markets.
Use a variety of methods to get a well-rounded picture of these proposed markets, the best
approach being a combination of qualitative and quantitative methods that provides picture
references, strength, beliefs, and anecdotes.
5. Look at the findings and analyze what must be done differently, abroad or internationally,
in comparison with current domestic marketing activities.5
Thus, while conducting IMR, one should be aware of the complicated cultural differences in differing regions of the world. This complication stems from operating in different and diverse environmental contexts, ranging from the technologically advanced and stable United States to mature
Western European markets, to the fast-changing environments in newly industrialized countries
such as Hong Kong (part of China) and South Korea, to developing economies such as India and
Brazil, to transforming economies such as the former Soviet Union and Eastern Europe, and to less
developed countries on the African continent. Marketing Research in Action 3-5 illustrates the differences between the United States and Canada.
MARKETING RESEARCH IN ACTION 3-5
Are Canadians Our Cousins Up North?
The United States and Canada are not only geographically neighbors, but are also each other’s largest
trading partner. FTA, NAFTA, and interfirm trade explain most of the increasing trade between the two
countries. Nearly 90 percent of the total Canadian population lives within 100 miles of the U.S. border.
So, can a U.S.-based manufacturer treat the Canadian market as a mere extension of the domestic markets? Absolutely not! Despite their proximity and close ties, important differences exist between the two
countries. These include demographic, economic, and cultural differences. For starters, as of June 2008,
the Canadian population is about 33,212,696 (Canada Stats), or roughly about 11 percent of the U.S. population 303,824,640 (U.S. Census) for the same period. Interestingly, over 65 percent of the population is
accounted for by two provinces—Ontario and Quebec. Ontario has a socialist government, which is prolabor and imposes more restrictions on business than the provincial governments or the United States.
The costs of doing business are, in general, higher in Canada. Bilingual labeling is required. Personal and
corporate income taxes are higher, as are transportation and distribution costs and interest rates. Apart
from these macro-level differences, there are differences in the way business is done. For example, offering a sales promotion, like a contest, entails special legal requirements that may vary from province to
province.
The cultural differences are as important as the demographic, economic, political, and legal differences. More than 80 percent of the Quebec population uses French as its first language, and Quebec
nationals are committed to sovereignty status. Canadians are extremely sensitive to environmental issues,
and even municipal governments have strict local environmental ordinances. To make matters even more
difficult, Canada uses the metric system!
SOURCE: Updated and adapted from “Do Your Homework Before You Start Marketing in Canada,” Marketing News, September 14, 1992.
59
CHAPTER 3 THE MARKETING RESEARCH PROCESS
Framing Research Questions in an International
Environment
Problems may not always be couched in the same terms in different countries or cultural contexts. This may be due to differences in socioeconomic conditions, levels of economic development, or differences in any of the macroenvironmental factors.
Several academic scholars have identified and have pointed out the major reason for the failure of businesses and marketing research projects in a foreign environment. The result has been
the self-reference criterion (SRC) adopted by researchers in defining the problem in a foreign country. SRC assumes that the environmental variables (cultural and others) that are prevalent in the
researcher’s domestic market are also applicable to the foreign country. This is a major cause for
the failure of research projects, since defining the problem is the most crucial step in the marketing research process.
One of the most frequent objectives of international marketing research is foreign market
opportunity analysis.6 Marketing Research in Action 3-6 is a good example of international
MARKETING RESEARCH IN ACTION 3-6
Pillsbury Presses Flour Power in India
The Pillsbury Doughboy has landed in India to pitch a product that he had just abandoned in America:
plain old flour. Pillsbury’s recent products sold in the United States have diversified from plain flour to
frozen cookie dough and microwave pizzas. In India, however, selling packaged flour is almost revolutionary. This is due to the fact that most Indian housewives still buy raw wheat in bulk. They then clean
it by hand, store it in huge metal hampers, and regularly carry some to a neighborhood mill, where it
is ground between two stones.
Pillsbury is onto a potentially huge business. India consumes about 74 million tons of wheat a year
(July 2001 est.), second only to China. (The United States consumes about 33 million tons, U.S. Census
July 2001 est.) Much of India’s wheat ends up as roti, a flat bread prepared on a griddle that accompanies almost every meal. In a nation where people traditionally eat with their hands, roti is the spoon.
Nevertheless, less than 1 percent of all whole wheat flour is sold prepackaged. India’s climatic extremes
and bad road conditions make it difficult to maintain freshness from mill to warehouse, let alone on
store shelves.
Starting a flour operation meant turning back the clock for Pillsbury. Though it was born as a U.S. flourmilling company 130 years ago, it exited from that business in the early 1990s to focus on products such as
frozen baked goods and ice cream. Pillsbury thought of introducing high-value products when it first
explored India. But it quickly learned that most Indians don’t have enough disposable income for such fare.
Pillsbury is betting that flour will generate enough sales volume to compensate for the razor-thin margin.
Marketing managers climbed into the attics where housewives store their wheat and accompanied them to
their tiny neighborhood flour mills. Pillsbury had hoped to establish contracts with existing mills, but inspectors found hygiene and safety at some properties to be compromising.
Many focus groups and lab tests later, Pillsbury came up with its packaged wheat blend, Pillsbury
Chaki Fresh Atta. Responding to consumers’ biggest concern, Pillsbury pitches the flour with a promise
that rotis made from it will stay soft “for 6 hours.” The company declines to say what ingredients keep
the flour tasting fresh, though it says there are no artificial preservatives. The packaging is made of a
robust plastic laminate that costs about two and a half times as much as the paper wrappers typically
used in the United States.
According to recent studies, the size of the Indian food industry is estimated at $69.4 billion, of which
the processed food industry accounts for $22 billion. The total Indian market for food-processing equipment amounted to $1.2 billion in 1999–2000 and is estimated to increase to over $2 billion by 2005, with
a predicted average annual growth rate of 15 to 18 percent over the next two years. The market is still
minuscule, and gains will largely depend on how quickly Indian housewives embrace convenience. Several local companies familiar with Indian tastes have launched branded flour in recent years, only to
flounder.
SOURCE: Updated from article by Miriam Jordan, The Wall Street Journal, May 5, 1999, p. B1.
60
THE INTERNATIONAL MARKETING RESEARCH PROCESS
marketing research. When a firm launches international activities, information can be accumulated to provide basic guidelines. The aim is not to conduct a painstaking and detailed
analysis of the world, but to gather information on questions that will help management narrow the possibilities for international marketing activities. Possible questions an international
marketing researcher might ask to achieve this objective include the following:
■ Do opportunities exist in foreign markets for the firm’s products and services?
■ Which foreign markets warrant detailed investigation?
■ What are the major economic, political, legal, and other environmental facts and trends in each
of the potential countries?
■ What mode of entry does the company plan to adopt to enter the foreign market?
■ What is the market potential in these countries?
■ Who are the firm’s present and potential customers abroad?
■ What is the nature of competition in the foreign markets?
■ What kind of marketing strategy should the firm adopt?
Marketing Research in Action 3-7 describes how the environment affects marketing a product,
thereby emphasizing the need for market research before product introduction.
MARKETING RESEARCH IN ACTION 3-7
Two of a Kind? Environmental Differences between
the United States and the United Kingdom
Lewis W. Griptight, Co., a baby products company based in the United Kingdom, experienced success
with its new Kiddiwinks line of infant and toddler accessories and decided to introduce it to their counterpart, Binky-Griptight division, in the United States. Since American mothers prefer their kids to be
well-groomed, a grooming set was added to the U.S. line.
Both U.S. and U.K. customers agree on the importance of safety and reject glass bottle products,
despite their environmentally friendly and hygienic features. Despite the similarities, extensive marketing research discovered some subtle environmental differences between the United States and the U.K.
These differences (cultural and legal), made a huge impact on the way the product was introduced in
the United States (and may trigger major changes in the British line).
One cultural difference concerned naming the product. Kiddiwinks, the British line, means “children”
to consumers in Britain, but is meaningless to U.S. consumers. To prepare the product for the U.S. market, the company named the U.S. line “Binkykids” because of the consumers’ high recognition and association of the Binky name with pacifiers.
Another difference dealt with cultural perception of kids’ products and led to dramatic changes in
color and packaging graphics. Conservative British mothers view kids’ items as medical products and
therefore prefer the symbolic white packaging of Kiddiwinks products. The Binkykids package, on the
other hand, was given a more stylish look and “prominently features the line’s signature green and lilac
colors.” Although American moms are very safety conscious, they still favor a more “flamboyant” or a
“cute and cuddly” look.
A third difference related to safety regulatory matters. U.S. consumers like innovative products and
Binky-Griptight introduced a pacifier, their flagship product, that plays lullabies when sucked on. However, this product would require redesign in the U.K. because of safety regulations.
For more interesting articles on marketing research go to http://www.marketingpower.com/ResourceLibrary/
Pages/default.aspx
SOURCE: Cyndee Miller, “Kiddi Just Fine in the U.K., but Here It’s Binky,” Marketing News, August 28, 1995, p. 8.
61
Summary
The research process consists of a series of stages or steps that guide the research project from conception
through to final recommendations. An overview of the domestic marketing research and the international
marketing research processes was presented. This chapter discussed the research purpose and the research
objective in detail. Chapter 4 will provide a discussion of the research design and implementation stages.
Together, the two chapters will provide a structure for the rest of the book.
The specification of the research purpose involves, first, the identification of the decision involved, its
alternatives, and the importance of its timing. Sometimes the decision is as general as: “Should our marketing program be changed?” In such cases it is also useful to specify the problem or opportunity that is
motivating the research, or the environmental surveillance objective. The purpose statement also should
consider who the research users are. There are times when identifying the research users and understanding their decisions and motives can significantly improve the effectiveness of the research.
The research objective involves the identification of the research questions. The answer to an appropriate research question should be relevant to the research purpose, and the question should be as specific
as possible. In particular, hypotheses should be developed whenever possible. The research boundaries
specification is also part of the research objective statement.
Even at the early stages of research conceptualization, it is useful to consider what value the resulting
information is likely to have. This exercise may lead to a decision to forgo the research, or at least make a judgment about the appropriate scale of the research project.
Questions and Problems
1. Jim Mitchell, a high-profile businessperson, is considering running for state governor against a twoterm incumbent. Mitchell and his backers do not
want to enter the race unless there is a reasonable
chance of winning. What are some research questions and hypotheses that, if answered, could help
him make the decision?
2. At the beginning of Chapter 1 there are two examples of management information needs: Esporta
Health and Fitness and Samsung Life Insurance.
Review each of these situations and develop an
appropriate set of research purposes and objectives.
3. In the United Kingdom, cars are polished more
frequently when the owners do not have garages.
Is the lack of a garage a good variable for predicting sales of car polish? Are there other hypotheses
that might explain this finding?
4. Can you think of additional research objectives for
the HMO study?
5. You have been retained by a manufacturer of major
appliances to investigate the probable color preferences for stoves and refrigerators in the coming
year. What is the purpose of the research? Are there
different purposes that might require different
research approaches?
6. The president of a small chain of women’s clothing stores was concerned about a four-year trend
of decreasing profits. The stores have been characterized as being rather conservative over the years
with respect to their product line, store decor, and
advertising. They have consistently avoided trendy
62
clothes, for example. Their market is now becoming extremely competitive because several aggressive fashion stores are expanding and are aiming
at the young, fashion-conscious buyer. As a result
of this competition and the disappointing profit
trend, the president is considering making the
product line appear less conservative and more
oriented toward the young buyer. Before making
such a risky change, the president feels it prudent
to conduct some marketing research to learn the
exact status of his chain. What should be the
research purpose? Compose a set of research questions that would be helpful.
7. Explain the role of an information system. How
would the possible role differ for Avon Products
and Johnson Wax Company? (Refer to the Agree
Shampoo example in Chapter 1 for background.)
8. Consider the example in Figure 3-4 in the Appendix to this chapter. What would be the expected
value of perfect information if the loss would be
$1 million instead of $2.5 million? How about if
the loss would be $10 million instead of $2.5 million? What would it be if the probability of failure
would be .2 instead of .4? Explain in words what
is meant by the expected value of perfect information and what its implication is.
9. Consider the example in Figure 3-4 in the Appendix
to this chapter. Determine the value of research information under the following situations:
a. Pr(Neg 兩 F) ⫽ .9 and Pr(Pos 兩 F) ⫽ .1 and all else
remains the same.
b. Pr(Pos 兩 S) ⫽ .7 and Pr(Neg 兩 S) ⫽ .3 and all else
remains the same.
c. Pr(Pos 兩 S) ⫽ .9, Pr(Neg 兩 S) ⫽ .1, Pr(Neg 兩 F) ⫽ .9,
Pr(Pos 兩 F) ⫽ .1.
10. ExoArt, Inc., a small U.S.-based manufacturer of
exotic jewelry, feels that a market exists for its
product in foreign markets. However, the company’s managers have no experience in the international environment, and do not know how to
proceed in forming a marketing strategy for international markets. They have decided to contact a
marketing research firm to help with the process.
The researchers recommend a “foreign market
opportunity analysis” as a starting point for the
company’s internationalization.
a. What is the aim of a foreign market opportunity
analysis?
b. Which questions might a researcher ask to
gather information for the analysis?
c. What is the most probable cause of failure for
a business or marketing research project in a
foreign environment?
11. a. Are there any differences between the basic
functions of marketing research in a domestic
environment and those in an international
environment?
b. Why is the international research process considered more complicated than the domestic
research process?
12. Crystal-Clear Lens, Inc., a newly formed mailorder contact lens company, has struggled to
obtain breakeven sales after five years in the eyewear market. The company’s founders felt that a
high demand would exist for mail-order supply as
a low-cost alternative to purchasing lenses at optical outlets. These retail outlets usually are within
close geographic proximity to an affiliated optician.
This allows customers to have their eyesight examined by the optician and then take the prescription
to the optical outlet to purchase their eyeglasses or
contact lens. Many retail outlets offer coupons that
refund the cost of the eye examination upon the
purchase of contact lenses, and offer several free
follow-up visits after the sale to check that the prescription has been made up correctly. The mailorder process requires customers to send in their
prescription after the eye examination, whereupon
the contact lenses will be supplied within two
weeks of receipt of the order. The managers at
Crystal-Clear Lens, Inc., have employed you as
their marketing research consultant, to determine
the reasons for the low sales.
a. What would be the research purpose of this
study?
b. How does the research purpose differ from the
research objective? Illustrate this difference in
terms of the Crystal-Clear Lens example.
c. What specific information would be required to
achieve the research purpose? (That is, state the
research question.)
d. State some preliminary hypotheses to answer
the research question.
End Notes
1. Lawrence D. Gibson, “Defining Marketing
Prolems—Don’t Spin Your Wheels Solving the
Wrong Puzzle,” Marketing Research, Spring 1998,
Vol. 10, No. 1, pp. 5–12.
2. Adapted from Irving D. Canton, “Do You Know
Who Your Customer Is?” Journal of Marketing, April
1976, p. 83.
3. “Conjoint Analysis of Values of Reserve Component Attitudes,” a report prepared for the
Department of Defense by Market Facts,
Chicago, November 1977.
4. Europa: http://europa.eu/
5. Allyson L. Stewart-Allen, “Do Your International
Homework First,” Marketing News, January 4, 1999,
Vol. 33, No. 1, p. 25.
6. Adapted from R. Michael Czinkota and Ilkka
A. Ronkainen, International Marketing, 3rd ed.,
Orlando, FL: Dryden Press, 1993.
Appendix
The Value of Research Information Using
Bayesian Decision Theory
In the discussion surrounding Figure 3-3, the value of
perfect research information for case B was determined
to be $1 million. Perfect research information would indicate with certainty that the product would be a success
or a failure. In this Appendix, the more realistic question
as to the value of imperfect research information will be
addressed.
The case B example of Figure 3-3 is shown at the top
of Figure 3-4. A decision to introduce a new product has
two possible outcomes. A “success” is given a prior
probability of .6 and will result in a profit of $4 million.
63
0.6
Introduce
S
EV = $1.4 million
0.4 F
$4 million
–$2.5 million
0.75
Introduce
.64 Pos
Concept
test
EV = $1.52 million
Introduce
Do not
introduce
EV = $0
Do not
introduce
$4 million
EV = $2.375 million
EV = $2.375 million
Do not
$0
introduce
.36 Neg
S
0.25
F
0.33
S
–$2.5 million
$4 million
EV = –$0.355 million
0.67
F
–$2.5 million
$0
$0
FIGURE 3-4
The concept test option.
A “failure” is given a prior probability of .4 and will
result in a profit of ⫺$2.5 million. Symbolically,
Product success ⫽ S ⫽ $4 million
Product failure ⫽ F ⫽ ⫺$2.5 million
Probability of success ⫽ Pr(S) ⫽ .6
Probability of failure ⫽ Pr(F) ⫽ .4
The expected value of introducing the product is then
the sum of the two possible payoffs, each weighted by
its probability of occurring:1
EV ⫽ expected value ⫽ S ⫻ Pr(S) ⫹ F ⫻ Pr(F)
⫽ ($4 million) ⫻ .6 ⫹ (⫺$2.5 million) ⫻ .4
⫽ $1.4 million
This expected value is shown in Figure 3-3 between the
two outcomes, success and failure. The alternative of
not introducing will result in a profit of zero. Thus, the
product will be introduced. If it were possible to obtain
perfect information about the outcome of the product
introduction, then the potential would be created to
avoid the result of losing $2.5 million. The maximum
amount that would be paid for perfect information
1
To understand expected value, consider the following game. Ten balls
numbered from 1 to 10 are placed in a bowl. One ball is selected at
random. If the number is 1, 2, 3, 4, 5, or 6, you receive $4 million. If it
64
would be $2.5 million times .4, the probability that a
failure will result, which amounts to $1 million. Only
when a failure results will the decision be affected;
therefore, the firm will save some money.
Figure 3-4 has added the alternative of evaluating
the new product with a concept test, such as a focus
group interview in which a group of 8 to 10 people is
asked to discuss the concept. Let us assume that the
concept test will either be positive (Pos) or negative
(Neg). Now assume that the concept test is not a perfect
indicator of the ultimate result. In particular, those who
have worked with this concept test several times before
believe that for successful products the concept test
is positive 80 percent of the time (and thus negative
20 percent of the time). For failed products, the concept
test is believed to be negative 60 percent of the time
(and thus positive 40 percent of the time). In terms of
conditional probabilities,
Probability of a Pos concept test
given a successful product
⫽ Pr1Pos|S2 ⫽ .8
Probability of a Neg concept test
given a successful product
⫽ Pr1Neg|S2 ⫽ .2
is 7, 8, 9, or 10, you must pay $2.5 million. If you played this game
thousands of times you would earn $1.4 million, on average, each time
the game was played.
Probability of a Neg concept test
given a failed product
⫽ Pr1Neg|F2 ⫽ .6
Probability of a Pos concept test
given a failed product
⫽ Pr1Pos|F2 ⫽ .4
The expected value of the “introduce the product
immediately” option is $1.4 million. We need to determine the expected value of the concept test alternative
in order to determine how much value, if any, there
will be in conducting the test. To proceed we first need
to determine
Pr(Pos) ⫽ probability of a positive concept test
Pr(Neg) ⫽ probability of a negative concept test
Pr(S 0 Pos) ⫽ probability that the product will be
successful given a positive concept test
Pr(F 0 Pos) ⫽ probability that the product will be a
failure given a positive concept test
Pr(S 0 Neg) ⫽ probability that the product will be
successful given a negative concept test
Pr(F 0 Neg) ⫽ probability that the product will be a
failure given a negative concept test
The first two terms are obtained from the following
formula:2
Pr(Pos) ⫽ Pr(Pos 0 S)Pr(S) ⫹ Pr(Pos 0 F)Pr(F)
⫽ .8 ⫻ .6 ⫹ .4 ⫻ .4 ⫽ .64
Pr(Neg) ⫽ Pr(Neg 0 S)Pr(S) ⫹ Pr(Neg 0 F)Pr(F)
⫽ .2 ⫻ .6 ⫹ .6 ⫻ .4 ⫽ .36
To obtain the balance of the expressions we apply
a formula known as Bayes’ theorem, which is the basis
of Bayesian decision theory:
Pr(S 0 Pos) ⫽
Pr(F 0 Pos) ⫽
Pr(S 0 Neg) ⫽
Pr(F 0 Neg) ⫽
Pr(Pos 0 S) ⫻ Pr(S)
.8 ⫻ .6
⫽
⫽ .75
Pr(Pos)
.46
Pr(Pos 0 F) ⫻ Pr(F)
Pr(Pos)
⫽
.4 ⫻ .4
⫽ .25
.64
Pr(Neg 0 S) ⫻ Pr(S)
⫽
Pr(Neg 0 F) ⫻ Pr(F)
.2 ⫻ .6
⫽ .33
.36
⫽
.6 ⫻ .4
⫽ .67
.36
Pr(Neg)
Pr(Neg)
The information is now at hand to evaluate the
concept test alternative. We first consider what happens
if there is a positive concept test and the product is
introduced. The probability of obtaining a successful
product, which would earn $4 million, is Pr(S | Pos) or
.75. The failure probability is of course Pr(F | Pos) or
.25. The expected value, the sum of the two outcomes
each weighted by its probability, is then
S ⫻ Pr(S 0 Pos) ⫹ F ⫻ Pr(F 0 Pos) ⫽ ($4 million)
⫻ .75 ⫹ (⫺$2.5 million) ⫻ .25 ⫽ $2.375 million
Since $2.375 million is more than the value of the “do not
introduce” alternative, the product would be introduced
after a positive concept test, and an expected value of
$2.375 million would result, as shown in Figure 3-4.
A similar analysis reveals that the expected value
of introducing the product after a negative concept test
would be
($4 million) ⫻ .33 ⫹ (⫺$2.5 million) ⫻ .67
⫽ ⫺$.355 million
Thus, given a negative concept test, the preferred alternative would be “do not introduce,” which would yield zero.
Therefore an expected value of $2.375 million
(associated with a positive concept test) will occur
with a probability of .64 [recall that Pr(Pos) ⫽ .6]. Further, an expected value of zero (associated with a negative concept test) will occur with probability .36. The
expected value of a concept test alternative is thus
($2.375 million) ⫻ .64 ⫹ ($0) ⫻ .36 ⫽ $1.52 million
Note that
Expected value of “concept test” alternative
⫽ $1,520,000
Expected value of “introduce” alternative
⫽ $1,400,000
Difference ⫽ $120,000
Thus the value of a concept test is $120,000, considerably
less than the value of perfect information ($1 million),
but still substantial. A researcher should be willing to
pay up to $120,000 for a concept test. Of course, if
another concept test could be found that would predict
success or failure more accurately, then its value would
be higher, as the reader could demonstrate.
2
The formulas are easily expanded if there are more than two outcomes to the decision. For example, suppose that there were three outcomes,
Success (S), Failure (F), and Indeterminant (I). The formula would then be
Pr1Pos2 ⫽ Pr1Pos 0 S2 ⫻ Pr1S2 ⫹ Pr1Pos 0 I2 ⫻ Pr1I2 ⫹ Pr1Pos 0 F2 ⫻ Pr1F2
65
CASE 3-1
A VideOcart Test for Bestway
Stores
The executives of Bestway Stores were intrigued with
a proposal they had received from Information
Resources, Inc., the developer of the VideOcart, to participate in a market test of the new point-of-sale technology. They were debating whether to agree to let IRI
conduct a test in 3 of their 300 stores. Their reasons for
doing it were not because they especially wanted to
help IRI, but rather to learn about the benefits and
shortcomings of this approach to in-store displays as a
possible competitive weapon. As an input to the decision, the marketing research department was asked to
design a study that would assess the desirability of
deploying the new display technology in all their stores
once the test was finished. The test was to last about
12 months.
Much of what the Bestway Stores’ management
knew about the VideOcart came from a press conference. They learned that IRI would beam a commercial
via satellite to a pick-up dish at each store. The signal
would be sent out by a low-power FM transmitter to
each cart, and stored in the memory of a computer
located in the handlebar of the cart.
According to an IRI spokesperson,
The ads will be shown at breaks in an information and
entertainment program for consumers and won’t interrupt the program. The sequence of the ads shown will
be determined by the route of the cart through the store.
As a shopper pushes a VideOcart down the aisles,
the manufacturer’s ads for brands on the shelves being
passed at that moment will be “triggered” at a rate of
about two per aisle (about 32 per store) and appear on
the flat, 6-in. by 8-in. liquid-crystal display mounted
on the handles of the cart. Tie-in promotion ads also
will be able to be used—for instance, a hotdog bun ad
when the cart is near the hot dogs.
Only about 15 percent of VideOcart’s display
time will be devoted to ads. The rest will be a
friendly medium in which to display ads, including
a continually changing video newsmagazine, news
to create a new shopping experience, store specials
and maps, trivia questions, and videogames to play
while waiting to check out.
The video seen on the screen isn’t television but
will use attention-getting graphics created on a personal computer.
How Does a VideOcart Work?
Here is what a consumer would find:
After a long day at work, you rush to the supermarket to pick up dinner. As you wheel a grocery
cart down the aisle, an ad for Brand X coffee flashes
on a liquid-crystal screen perched on your cart’s handlebar. The ad reminds you that you need coffee, so
you drop a can of Brand X into your cart and push
on to the next aisle.
BOB’S SUPERMARKET
Checkout
transmitter
Message
Store
transmitter
Manager's PC
66
The Test of the VideOcart
Besides gathering sales data, the test markets will be
used to perfect consumer programming and fine-tune
the technology. IRI also will be checking on factors such
as the ideal length of ads, shopper interest in games and
information, how the shopper interacts with the unit,
and opportunities for the grocers to contribute programming. VideOcart’s computer capabilities also will
offer supermarkets some advantages, including sounding an alarm if a cart is taken too far from the store, as
in an attempt to steal it. The stores also could use
VideOcart to transmit information such as the shortest
checkout line, the next number up at the deli counter,
or that a red Ford in the parking lot has its lights on.
We will put the needed equipment into the
supermarkets—including a satellite dish on the
roof—at no cost to the retailers and eventually will
pay the supermarkets a royalty four to six times
greater than they’re now receiving from other shopping cart ads. We’re not asking for an exclusive in
the supermarkets, but we’re out to make that other
type of shopping cart ad irrelevant.
The VideOcarts will be designed to be weatherproof and childproof. The retailers will have to
recharge the batteries on the carts each night.
The carts should be equipped for under $500 per
cart. The average supermarket has about 100 carts
and IRI would turn about 75 of them into VideOcarts, which we expect shoppers will seek out because
they’ll make shopping efficient and fun.
VideOcart will offer excellent media efficiency
compared to other alternatives. Once national, the
medium will reach 60 percent of all shoppers in a
week at the cost per thousand of a free-standing
newspaper insert—$4–$5 per thousand households.
VideOcart ads will be able to be created in a few
hours and somewhat inexpensively using microcomputer graphics software. The learning curve to master the technology should be short and IRI will provide technical support and counsel.
The Assignment
The marketing research manager pondered about this
project and wondered what really had to be learned to
draw up a statement of purpose. With this in hand the
design of the research would be a lot easier. Time was
short, for management needed to know how the
research manager was going to get usable information.
If it looked too difficult or expensive, they might not
agree to a test in their stores.
CASE 3-2
Sperry/MacLennan Architects
and Planners
year, the company was incorporated as H. Drew Sperry
and Associates; by then Sperry had added three junior
architects, a draftsman, and a secretary. One of those
architects was John MacLennan, who would later
become a senior partner in Sperry/MacLennan.
Throughout the 1970s, the practice grew rapidly
as the local economy expanded, even though the market
for architectural services was competitive. With
the baby-boom generation entering the housing market,
more than enough business came its way to enable
Sperry to develop a thriving architectural practice,
and by 1979 the company had grown to 15 employees
and had established branch offices in Charlottetown and
Fredericton. These branch offices had been established to
provide a local market presence and meet licensing
requirements during this aggressive growth period.
But the growth could not last. The early 1980s was
not an easy time for the industry, and many architectural firms found themselves unable to stay in business
through a very slow period in 1981–1982. The company
laid off all but the three remaining partners: Drew,
Sheila Sperry, and John MacLennan. However, one
draftsman and the secretary refused to leave, working
In August 1988 Mitch Brooks, a junior partner and
director of Sperry/MacLennan (S/M), a Dartmouth,
Nova Scotia, architectural practice specializing in recreational facilities, is in the process of developing a plan to
export his company’s services. He intends to present the
plan to the other directors at their meeting the first week
of October. The regional market for architectural services
is showing some signs of slowing, and S/M realizes that
it must seek new markets. As Sheila Sperry, the office
manager and one of the directors, said at their last
meeting, “You have to go wider than your own backyard.
After all, you can only build so many pools in your own
backyard.”
About the Company
Drew Sperry, one of the two senior partners in
Sperry/MacLennan, founded the company in 1972 as a
one-man architectural practice. At the end of its first
67
without pay for several months in the belief that the
company would win a design competition for an aquatics center in Saint John; their faith in the firm is still
appreciated today.
Their persistence and faith was rewarded in 1983.
Sperry won the competition for the aquatics facility for
the Canada Games to be held in Saint John. Sperry had
gained national recognition for its sports facility expertise, and its reputation as a good design firm specializing in sports facilities was secured.
From the beginning, the company found recreational facilities work to be fun and exciting. To quote
Sheila Sperry, this type of client “wants you to be innovative and new. It’s a dream for an architect because it
gives him an opportunity to use all the shapes and colors and natural light. It’s a very exciting medium to
work in.” So they decided to focus their promotional
efforts to get more of this type of work and consolidate
their “pool designer” image by associating with Creative
Aquatics on an exclusive basis in 1984. Creative Aquatics provided aquatics programming and technical operations expertise (materials, systems, water treatment,
safety, and so on) to complement the design and planning skills at Sperry.
The construction industry rebounded in 1984;
declining interest rates ushered in a mini-building
boom, which kept everyone busy for the 1984–1987
period. Mitch Brooks joined the practice in 1987. The
decision to add Brooks as a partner, albeit a junior one,
stemmed from their compatibility. Brooks was a good
production architect, and work under his supervision
came in on budget and on time, a factor compatible
with the Sperry/MacLennan emphasis on customer
service. The company’s fee revenue amounted to
approximately $1.2 million in the 1987 fiscal year; however, salaries are a major business expense, and profits
after taxes (but before employee bonuses) accounted
for only 4.5 percent of revenue.
Now it is late August, and with the weather cooling,
Mitch Brooks reflects on his newest task, planning for the
coming winter’s activities. The company’s reputation in
the Canadian sports facility market is secure. The company has completed or has in construction five sports
complexes in the Maritime Provinces and five in Ontario,
and three more facilities are in design. The awards have
followed, and just this morning, Drew was notified of
their latest achievement—the company has won the
$10,000 Canadian Architect Grand Award for the Grand
River Aquatics and Community Center near Kitchener,
Ontario. This award is a particularly prestigious one
because it is given by fellow architects in recognition of
design excellence. Last week Sheila Sperry received word
that the Amherst, N.S., YM-YWCA won the American
National Swimming Pool and Spa Gold Medal for pool
design against French and Mexican finalists, giving them
international recognition. Mitch Brooks is looking
forward to his task. The partners anticipate a slight
68
slowdown in late 1988, and economists are predicting a
recession for 1989. With 19 employees to keep busy and
a competitor on the West Coast, they decided this morning that it is time to consider exporting their hard-won
expertise.
The Architecture Industry
Architects are licensed provincially, and these licenses
are not readily transferable from province to province.
Various levels of reciprocity are in existence. For this
reason, joint ventures are not that uncommon in the
business. In order to cross provincial boundaries, architecture firms in one province often enter into a joint
venture arrangement with a local company.
It is imperative that the architect convince the
client that he or she has the necessary experience and
capability to undertake the project and to complete it
satisfactorily. S/M has found with its large projects that
the amount of time spent meeting with the client
requires some local presence, although the design need
not be done locally.
Architects get business in a number of ways.
“Walk-in” business is negligible, and most of S/M’s
contracts are the result of one of the following five
processes:
1.
2.
3.
4.
5.
A satisfied client gives a referral.
A juried design competition is announced. (S/M
has found that these prestigious jobs, even though
they offer “runners up” partial compensation, are
not worth entering except to win, since costs are
too high and the compensation offered other
entrants too low. Second place is the same as last
place. The Dartmouth Sportsplex and the Saint
John Aquatic Center were both design competition
wins.)
A client publishes a “Call for Proposals” or a “Call
for Expressions of Interest” as the start of a formal
selection process. (S/M rates these opportunities;
unless it has a 75 percent chance of winning the
contract, it views the effort as not worth the risk.)
A potential client invites a limited number of architectural firms to submit their qualifications as the
start of a formal selection process. (S/M has a prepared qualification package that it can customize
for a particular client.)
S/M hears of a potential building and contacts the
client, presenting its qualifications.
The fourth and fifth processes are the most common in buildings done for institutions and large corporations. Since the primary buyers of sports facilities
tend to be municipalities or educational institutions,
these are the ways S/M acquires a substantial share of
its work. Although juried competitions are not that
common, the publicity possible from success in landing
this work is important to S/M. The company has found
that its success in securing a contract is often dependent on the client’s criteria and the current state of the
local market, with no particular pattern evident for a
specific building type.
After the architect signs the contract, there will be
a number of meetings with the client as the concept
evolves and the drawings and specifications develop.
Therefore, continuing client contact is as much a part
of the service sold as the drawings, specifications, and
site supervision and, in fact, may be the key factor in
repeat business.
Developers in Nova Scotia often are not loyal buyers, changing architects with every major project or
two. Despite this, architects are inclined to think the
buyer’s loyalty is greater than it really is. Therefore,
S/M scrutinizes buyers carefully, interested in those
that can pay for a premium product. S/M’s philosophy
is to provide “quality products with quality service for
quality clients,” and thus produce facilities that will
reflect well on the company.
The Opportunity
In 1987, a report entitled “Precision, Planning, and
Perseverance: Exporting Architectural Services to the
United States” identified eight market niches for
Canadian architects in the United States, one of which
was educational facilities, in particular post-secondary
institutions. This niche, identified by Brooks as most
likely to match S/M’s capabilities, is controlled by state
governments and private organizations. Universities are
known not to be particularly loyal to local firms and so
present a potential market to be developed. The study
reported that “post-secondary institutions require design
and management competence, whatever the source”
(p. 39). Athletic facilities were identified as a possible
niche for architects with mixed-use-facility experience.
Finally, the study concluded that “there is an enormous
backlog of capital maintenance and new building requirements facing most higher education institutions” (p. 38).
In addition to the above factors, the study indicated others that Brooks felt were important:
1.
2.
3.
4.
The United States has 30 percent fewer architectural firms per capita than Canada.
The market shares many Canadian values and
work practices.
The population shift away from the Northeast to
the Sunbelt is beginning to reverse.
Americans are demanding better buildings.
Although Brooks knows that Canadian firms have
always had a good reputation internationally for the
quality of their buildings, he is concerned that American firms are well ahead of Canadian ones in their
use of CADD (computer-assisted design and drafting)
for everything from conceptual design to facility management. S/M, in spite of best intentions, has been
unable to get CADD off the ground, but is in the
process of applying to the Atlantic Canada Opportunities Agency for financial assistance in switching
over to CADD.
Under free trade, architects will be able to engage
freely in trade in services. Architects will be able to travel
to the United States and set up an architectural practice
without having to become qualified under the American
Institute of Architects; as long as they are members
of their respective provincial associations and have
passed provincial licensing exams and apprenticeship
requirements, they will be able to travel and work in the
United States and import staff as required.
Where to Start?
At a meeting in Halifax in January 1988, the Department
of External Affairs had indicated that trade to the United
States in architectural services was going to be one positive benefit of the Free Trade Agreement to come into
force in January 1989. As a response, S/M has targeted
New England for its expansion, because of its geographic
proximity to S/M’s homebase in the Halifax/Dartmouth
area and also because of its population density and similar climatic conditions. However, with all the hype about
free trade and the current focus on the United States,
Brooks is quite concerned that the company might be
overlooking some other very lucrative markets for his
company’s expertise. As part of his October presentation
to the board, he wants to identify and evaluate other
possible markets for S/M’s services. Other parts of the
United States, or the affluent countries of Europe, where
recreational facilities are regularly patronized and design
is taken seriously, might provide a better export market,
given S/M’s string of design successes at home and
the international recognition afforded by the Amherst
facility design award. Brooks feels that designing two
sports facilities a year in a new market would be an
acceptable goal.
In his search for leads, Brooks notes that the APPA
(Association of Physical Plant Administrators of Universities and Colleges) charges $575 for a membership,
which provides access to its membership list once a year.
But this is only one source of leads. And of course there
is the U.S. Department of Commerce, Bureau of the
Census, as another source of information for him to tap.
He wonders what other sources are possible.
S/M appears to have a very good opportunity in the
New England market because of all its small universities
and colleges. After a decade of cutbacks on spending,
corporate donations and alumni support for U.S. universities has never been so strong, and many campuses have
sports facilities that are outdated and have been poorly
maintained. But Mitch Brooks is not sure that the New
England market is the best. After all, a seminar on exporting that he attended last week indicated that the most
geographically close market, or even the most psychically
close one, may not be the best choice for long-run profit
maximization and/or market share.
69
Questions for Discussion
1. What types of information will Brooks need to collect before he can even begin to assess the New
England market? Develop a series of questions you
feel are critical to this assessment.
2. What selection criteria do you believe will be relevant to the assessment of any alternative markets?
What preliminary market parameters are relevant
to the evaluation of S/M’s global options?
CASE 3-3
Philip Morris Enters Turkey
Under regulatory and legal fire in the United States,
tobacco companies have been staking their financial
futures in developing countries. In the seven years
since Turkey abolished price controls that had propped
up its state-owned tobacco company, Philip Morris has
gotten millions of Turks to cast aside local cigarettes in
favor of its Marlboro, Parliament, and L&M brands.
For Philip Morris—which saw international tobacco
profits grow 60 percent to $4.6 billion last year from
1994—nowhere has that push been more successful
than in Turkey. In the nation that inspired the phrase
“smokes like a Turk,” 43 percent of the 62.9 million
population smokes, according to government estimates, compared to 25 percent in the United States.
Meantime, cigarette consumption in Turkey has
increased at an annual rate of about 4.76 percent since
1992, making this one of the fastest-growing markets
in the world. Philip Morris has watched its share of
the Turkish cigarette market mushroom, to 23 percent
in 1997 from 15 percent in 1995, while Tekel, the
government-owned tobacco company, has seen its
hold drop to 70 percent from 82 percent over the same
period, according to figures provided by the Turkish
government. Third-ranked R.J. Reynolds, with its
Winston and Camel brands, says its share has grown
to 7.3 percent from 2.9 percent.
Turkish smokers got their first taste of Philip
Morris brands in the 1970s, when smuggled, tax-free
American cigarettes began flooding local bazaars. But
it wasn’t until the early 1980s that the company was
allowed to sell cigarettes in Turkey. Eager to raise
Turkey’s status in the West, then–Prime Minister Turgut
Ozal decided to turn his rural country into a model of
free enterprise. One of his targets: Tekel, the creaky
state monopoly that has held exclusive rights to sell
tobacco, salt, and liquor to Turks since the waning days
of the Ottoman Empire.
How could Mr. Ozal prod Tekel into the modern
age? He announced in 1984 that foreign tobacco
70
3. Assuming that S/M decides on the New England
market, what information will be needed to implement an entry strategy?
Source: This case was adapted with permission from
a case prepared by Dr. Mary R. Brooks, of Dalhousie
University.
merchants would be allowed in Turkey for the first
time since the days of the sultans.
But there was a catch: Tekel would continue
to price and distribute all cigarettes both foreign
and domestic. That advantage came in especially
handy when Tekel in 1988 launched Tekel 2000,
a cigarette blended with American tobacco leaves that
it designed to compete with Marlboro. Priced about
25 cents lower than a standard 20-cigarette pack
of Marlboros at the time, Tekel 2000 quickly won a
quarter of the market.
Arguing that it couldn’t survive in Turkey unless
it had the right to price and distribute its own products, Philip Morris leveraged the one thing it had that
the government badly wanted: millions of dollars to
invest in the country. And it wouldn’t invest that
money unless Tekel gave up control.
Tekel eventually relented, and in May 1991,
Philip Morris got the right to market, price, and distribute its own cigarettes, conditioned on a number of
factors, including building its own factory. Philip
Morris announced a joint venture with Sabanci Holding
Inc., a local company, and poured $100 million to
start construction of a factory in the southwestern city
of Torballi. The factory opened in 1993; Philip Morris
eventually expanded it into a $230 million facility
capable of cranking out more than 28 billion cigarettes annually.
Supporting the idea that Philip Morris adjust its
cigarette blends is a confidential 1992 report, titled “PM’s
Global Strategy: Marlboro Product Technology,” conducted by researchers at rival B.A.T. Industries PLC’s
Brown & Williamson unit. “When Marlboro has been
introduced into a market there is no evidence that initial offerings may be closer to that market’s traditional
taste,” concludes the report that documents different
Marlboro formulations in countries including Brazil,
Britain, and Germany. “Over time PM will alter the
product and introduce product technology more consistent with an overall Marlboro sensory character.”
Philip Morris says it doesn’t comment on speculation by competitors. The company says it strives “to
ensure that Marlboros are as consistent as possible
worldwide,” but adds that some variation results from
local regulations that limit constituents like tar or
require the use of locally grown tobaccos. Beyond the
taste difference, Tekel couldn’t keep up when Marl-
boro’s prices were cut. Although they still maintain
Tekel 2000 was a success, Tekel executives announced
that they want to sell the brand to a foreign competitor. Where Philip Morris really bested rivals is in marketing and distribution.
Questions for Discussion
1. Obtain information on the current consumption
pattern of cigarettes in Turkey. How has this
changed in the last 10 years? What kind of research
should Philip Morris be doing to gain market share
in Turkey?
2. What kind of information should Philip Morris
obtain to decide on a marketing and distribution
strategy in Turkey? How can Philip Morris obtain
this information?
3. Is it ethical for Philip Morris to overtly alter the
taste of its cigarettes over a period of time for its
own profit?
Source: This case was prepared by V. Kumar and
Rajkumar Venkatesan for the purpose of classroom
discussion from “How Philip Morris Got Turkey
Hooked on Marlboro,” Wall Street Journal,
September 1998.
71
4
RESEARCH DESIGN AND
IMPLEMENTATION
Learning Objectives
■ Understand the definition and purpose of research design.
■ Be familiar with the different types of research designs.
■ Identify the appropriate data collection method for a given research design.
■ Describe and briefly discuss the various sources of errors in a design.
■ Be introduced to the concepts of budgeting and scheduling a project.
■ Describe the elements of a research proposal.
■ Be introduced to the issues in international marketing research design.
A research design is the detailed blueprint used to guide a research study toward its objectives.
The process of designing a research study involves many interrelated decisions. The most
significant decision is the choice of research approach, because it determines how the information will be obtained. The chosen research should be flexible enough to accommodate decisionmaking systems that, in turn, would contribute to valuable marketing intelligence tools. Typical
questions at this stage are: Should we rely on secondary sources such as the Census? Which is
more appropriate, an exploratory approach with group discussions or a survey? Is a mail, telephone, fax, or personal interview survey better for this problem?
Tactical research decisions are made once the research approach has been chosen. Here the
focus is on the specific measurements to be made or questions to be asked, the structure and
length of the questionnaire, and the procedure for choosing a sample to be interviewed. These
tactical decisions also are constrained by time and budget availability, so before a study can be
implemented, the estimated costs must be compared to the anticipated value.
To design something also means to ensure that the pieces fit together. The achievement of
this fit among objectives, research approach, and research tactics is inherently an iterative process
in which earlier decisions are constantly reconsidered in light of subsequent decisions. This may
mean a revision of the research objectives as new insights are gained into the complexities of the
population to be sampled, or a reassessment of the research approach in light of realistic cost
estimates. Consequently, few researchers find they have designed their research studies in the
neat and linear fashion that is implied by Figure 4-1; however, this figure is a useful overview
of major research design topics to be introduced in this chapter. Also in this chapter we will discuss the research proposal as a vehicle for summarizing significant decisions made during the
research design process.
72
Planning system
Information system
Preliminary planning stage
Research Purpose
Decision alternatives
Problem or opportunity
Research users
Research Objective
Research question
Hypothesis development
Research boundaries
Estimate Value of
Research Information
Research Approach
Research design
Exploratory
Descriptive
Causal
Choice of data collection method
• Secondary and standardized data
• Qualitative methods
• Surveys
• Experiments
Role of research supplier
• Project design
• Raw data collection
Research Tactics
Develop measures of interest
Construct questionnaire
Design experiments
Design sampling plan
Anticipate analysis
Compare Cost and Timing Estimates
with Anticipated Value
Terminate
Revise
Implementation
Proceed
Data Collection and Analysis
Data collection
Field work
Data processing
Data analysis
Statistical analysis
Interpretation
Conclusions and Recommendations
FIGURE 4-1
The research design process.
73
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
RESEARCH APPROACH
The choice of a research approach depends on the nature of the research that one wants to do.
In this section the various types of research approaches, data collection methods, and the factors
affecting their choice are discussed.
Types of Research
All research approaches can be classified into one of three general categories of research:
exploratory, descriptive, and causal. These categories differ significantly in terms of research purpose, research questions, the precision of the hypotheses that are formed, and the data collection
methods that are used.
Exploratory Research. Exploratory research is used when one is seeking insights into the general nature of a problem, the possible decision alternatives, and relevant variables that need to be
considered. Typically, there is little prior knowledge on which to build. The research methods are
highly flexible, unstructured, and qualitative, for the researcher begins without firm preconceptions as to what will be found. The absence of structure permits a thorough pursuit of interesting ideas and clues about the problem situation.
Exploratory research hypotheses are either vague and ill defined, or they do not exist at all.
Table 4-1 illustrates this point with three examples. In the first example, the research question asks
TABLE 4-1
Three Research Approaches
Research Purpose
Exploratory research
1. What new product should
be developed?
Research Question
Hypothesis
What alternative ways are there to
provide lunches for school children?
Boxed lunches are better than other
forms.
What benefits do people seek from
the product?
Constructs unknown.
What is the nature of any
customer dissatisfaction?
Suspect that an image of impersonalization is a problem.
Descriptive research
4. How should a new product be distributed?
Where do people now buy similar
products?
Upper-class buyers use specialty stores,
and middle-class buyers use department
stores.
5. What should be the
target segment?
What kinds of people now buy the
product, and who buys our brand?
Older people buy our brand, whereas
the young married are heavy users of
competitors’.
6. How should our product
be changed?
What is our current image?
We are regarded as being conservative
and behind the times.
What is the relationship between
size of service staff and revenue?
For small organizations, an increase of
50 percent or less will generate marginal
revenue in excess of marginal costs.
Advertising program A generates more
new riders than program B.
2. What product appeal will
be effective in
advertising?
3. How can our service be
improved?
Causal research
7. Will an increase in the
service staff be
profitable?
8. Which advertising
program for public
transit should be run?
What would get people out of cars
and into public transit?
9. Should a new budget or
“no frills” class of airfare
be introduced?
Will the “no frills” airfare generate
sufficient new passengers to offset
the loss of revenue from existing
passengers who switch from
economy class?
74
The new airfare will attract sufficient
revenue from new passengers.
RESEARCH APPROACH
MARKETING RESEARCH IN ACTION 4-1
Online Research Communities
Traditional marketing research is of late increasingly spilling over into cheaper and noninvasive media
such as the Internet. Companies such as Kraft Foods and Coca-Cola have created private online
research communities to collect data on consumer shopping behavior. Other companies like Dell and
MarketFoods have “democratized” innovation through active interaction and participation of consumers in online forums. These online communities are increasingly being viewed as collective and
outsourced research and development hubs where consumer problems surface and innovative solutions are brainstormed.
what alternative ways there are to provide lunches for school children. It was precipitated by information suggesting problems with existing school lunch programs. In this case, there simply is no
information that would suggest even the most tentative of hypotheses. In the second example, the
research question is to determine what benefits people seek from the product. Since no previous
research considered consumer benefits, it is difficult even to provide a list of them. In the third
example, the hypothesis is advanced that a root cause of customer dissatisfaction is an image of
impersonalization. However, this hypothesis is extremely tentative and provides at best only a partial answer to the research question.
Exploratory research is also useful for establishing priorities among research questions and
for learning about the practical problems of carrying out the research. Marketing Research in
Action 4-1 illustrates how various companies are using online surveys to conduct exploratory
research to gain insight into consumers’ lifestyles.
A variety of productive exploratory approaches will be discussed in Chapters 5 through 8,
including literature reviews, individual and group unstructured interviews, and case studies.
Descriptive Research. Descriptive research embraces a large proportion of marketing
research. The purpose is to provide an accurate snapshot of some aspect of the market environment, such as:
■ The proportion of the adult population that supports the United Way.
■ Consumer evaluation of the attributes of our product versus competing products.
■ The socioeconomic and demographic characteristics of the readership of a magazine.
■ The proportion of all possible outlets that are carrying, displaying, or merchandising our products.
In descriptive research, hypotheses often will exist, but they may be tentative and speculative.
In general, the relationships studied will not be causal in nature. However, they may still have
utility in prediction.
In the fourth example in Table 4-1, the research question concerns where people buy a particular type of product. One hypothesis is that upper-class families buy this type of product in
specialty stores and middle-class families use department stores. There is no explicit cause–effect
relationship. The question is simply to describe where people buy. With this hypothesis it is
clear that if data are gathered, it will be important to include indicators of social class and to
be prepared to analyze the data with respect to stores classified as specialty and department
stores. Thus, the development of the hypothesis provides guidance to the researcher by introducing more detail to the research question. Similarly, in the sixth example, the hypothesis suggests that when image is being measured, it is necessary to include measures of innovativeness.
Marketing Research in Action 4-2 provides an example of the outcome of descriptive research.
Causal Research. When it is necessary to show that one variable causes or determines the values of other variables, a causal research approach must be used. Descriptive research is not sufficient, for all it can show is that two variables are related or associated. Of course, evidence of
a relationship or an association is useful; otherwise, we would have no basis for even inferring
75
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
MARKETING RESEARCH IN ACTION 4-2
Does Color Matter?
In the soft-drink industry, Coke gave rise to Cherry Coke and Diet Coke. If flavor makes a difference
in the drink, can colors do the same for cars? Research experts and color experts strongly agree. For
new-car buyers, color and appearance are nearly as important as price. Automakers need to predict
which color will be “hot” in the future and change 30 percent of their color offerings based solely on
sales trends and the opinions of color experts. Since color is a subjective issue, consumers can’t foretell their color choice far in advance of a purchase decision and tend to be trend-conscious.
Cooper Marketing Group conducts a nationwide consumer color preference study annually and
devised a “color lifestyle” segmentation to identify the “movers and shakers in color.” The group
designed a series of psychographic statements for respondents to choose from, which concerned the
influence of color on purchasing decisions, and came up with three clearly defined color personalities:
■ The color-forward consumer likes to be the first to try a “new” color and is willing to spend more for
a product in a fashionable color. Color-forward people tend to be women under 30 or over 50, or
men under 30; city dwellers; impulse buyers; people who make less than $35,000 per year.
■ Color-prudent consumers will buy a “new” color only after seeing friends try it. They often put quality ahead of color when choosing products. Color-prudent people tend to be men or women age 30
to 50; suburban; careful shoppers; people who make more than $50,000 per year.
■ Color-loyal consumers replace a product with another of the same color and prefer safe colors such as
blue or gray rather than fashionable colors. Color-loyal people tend to be men over 60; suburban or
rural; people who dislike shopping; and may fall anywhere on the income spectrum.
The color-forward group represents a small, but highly influential, consumer segment. Color-prudent
shoppers make up the bulk of the market. Color-loyalists are a small, highly predictable segment because
they repeatedly buy the same color. This demographic data allows automakers to examine reasons
behind color trends and is useful in advertising and product positioning.
Car manufacturers try to use color to establish brand character and distinguish themselves from their
cousins or competition. Colors must not only be designed with consumer preferences in mind, they must
also take into account the market segment and physical characteristics of each car. For example, “muscle”
cars of the 1960s had colors that even Crayola’s crayons don’t include—”hemmy orange,” “top banana,”
and “plum crazy”—and are a perfect example of the value color can bring to the marketing identity of
a vehicle.
Consumers rank green as the hottest shade (19.4 percent), followed by white (18.1 percent), light
brown (11.8 percent), medium red (10 percent), and black (5.7 percent) as researched by Dupont Automotive Company. According to Cooper Marketing group, consulting consumers on color preferences
before deciding on “hot” colors is like taking 5,000 consumers into a boardroom meeting. This research
can effectively counter influential executives who simply decide against a recommended color.
For more interesting articles on marketing research go to http://www.marketingpower.com/ResourceLibrary/Pages/
default.aspx
SOURCE: Adapted from Tim Triplett, “Carmakers Driven by Quest to Find Tomorrow’s Color,” Marketing News, August 28, 1995, p. 38; Tim Triplett, “Research
Probes How Consumers Rely on Color for Their Purchases,” Marketing News, July 1, 1996, pp. 1, 39.
76
that causality might be present. To go beyond this inference we must have reasonable proof that
one variable preceded the other and that there were no other causal factors that could have
accounted for the relationship.
Suppose we had evidence that territories with extensive sales coverage, as measured by the
number of accounts per salesperson, had higher per-capita sales. Are there sufficient grounds for
a decision to increase sales coverage in areas where sales currently are weak? The answer would
depend first on whether past increases in sales coverage had led to increases in sales. Perhaps
the allocation of the sales force annual budget was based on the previous year’s sales. Then we
might conclude that past sales increases led to an increase in sales coverage—a conclusion with
dramatically different implications. Second, we would have to be sure that there were no other
reasons for differences in sales between territories. Perhaps the weak sales territories had special
RESEARCH APPROACH
MARKETING RESEARCH IN ACTION 4-3
Is Everyday Low Pricing Leading to Everyday Low Profits?
Over the past six months, Procter & Gamble has announced that 50 percent of its volume was on “value
pricing”—its name for everyday low prices—and that it expects to save $175 million from the shift. But
a new report from Salomon Brothers offers a more sobering picture of P&G’s trendline at the checkout
stand. In 10 of 11 household product categories it tracked, P&G’s dollar market share in supermarkets
fell in 1992. The million-dollar question now facing the retail industry is, “Should supermarkets adopt
the everyday low prices (EDLP) strategy? Does the EDLP strategy provide greater profits over traditional pricing strategies?”
A group of researchers from the University of Chicago conducted an experiment to find answers to
these questions. The researchers manipulated prices in 19 product categories in 88 stores of Dominick’s
Finer Foods, Inc., based in Chicago, and patronized by an estimated 1 million people each week. Some
stores used the standard pricing approach, called “high–low” in the industry. Others were converted to
everyday low pricing, in which prices were reduced and kept low. In their analysis of everyday low
pricing, the researchers moved prices up and down 10 percent in the key categories—which included
beer, cereals, cigarettes, detergents, frozen entrees, juices, and soft drinks—and accounted for 30 percent
of an average store’s sales. In stores with everyday low pricing in those categories, prices were dropped
an additional 10 percent. In stores with a high–low strategy, prices were raised 10 percent to test consumer response.
The result: Stores featuring everyday low pricing rang up slightly more sales but much less profit
than the high–low stores. Overall, profits in the categories that used everyday low pricing were about
17 percent below what grocers would have made with the traditional high–low approach, the researchers
calculate. They attribute the difference to the higher profit margins on items that aren’t on sale.
To find out more about other Procter & Gamble products go to http://www.pg.com.
SOURCE: Adapted from Jon Berry, “So How Is P&G’s Share? Lagging, New Study Says,” Brandweek, April 19, 1993, p. 16; Richard Gibson, “Broad
Grocery Price Cuts May Not Pay,” Wall Street Journal, May 7, 1993, pp. B1, B8.
requirements because of climate differences, and our product was at a disadvantage; or perhaps
the weak territories were served by competitors with local advantages. In either case, adding
more salespeople to weak sales territories would not improve sales, for the basic problems still
would be present.
Because the requirements for proof of causality are so demanding, the research questions and
relevant hypotheses are very specific. The examples in Table 4-1 show the level of detail that is
desirable. Marketing Research in Action 4-3 describes an application of causal research.
Detective Funnel
Each of the three types of research—exploratory, descriptive, and causal—has a distinct and complementary role to play in many research studies. This is most evident in studies that are initiated with this question: Why are our sales (share, patronage, contributions) below our objectives
or below last year’s performance? The first step is to use exploratory techniques to generate all
possible reasons for the problem (as shown in the drawing). Thereafter, a combination of descriptive and causal approaches is used to narrow the possible causes. Hence, the research is used in
exactly the same way that a detective proceeds to eliminate unlikely suspects. Descriptive
research evidence is often sufficient to filter out many of the possible causes.
For example, a municipal transit company, seeking to understand why ridership has
declined suddenly, can quickly dispose of weather-related factors by examining weather
records to see whether the recent weather pattern has been unusual. Similarly, evidence from
customer records can be used to determine whether or not telephone complaints about the
quality of service have increased. Also, their surveys of customers will reveal that service frequency and fares are the two most important factors in evaluating transit service, whereas riders are indifferent to the amount and type of advertising inside buses. If fares have not risen
or the costs of competitive transportation modes such as car parking or operating costs have
77
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
not declined, then attention can be focused on service frequency. Whether this is the causal
factor depends on whether there was a reduction in frequency that preceded the decline in
ridership.
Descriptive
research
Problem
Possible
causes
of the
problem
Exploratory
research
Probable
causes
Causal
research
Data Collection Methods
The research designer has a wide variety of methods to consider, either singly or in combination. They can be grouped first according to whether they use secondary or primary sources of
data. Secondary data are already available, because they were collected for some purpose other
than solving the present problem. Included here are (1) the existing company information system; (2) databanks of other organizations, including government sources such as the Census
Bureau or trade association studies and reports; and (3) syndicated data sources, such as consumer purchase panels, where one organization collects reasonably standardized data for use by
client companies. These secondary sources are discussed in Chapters 5 and 6. Primary data are
collected especially to address a specific research objective. A variety of methods, ranging from
qualitative research to surveys to experiments, may be employed. These methods are described
in more detail in Table 4-2. Some methods are better suited to one category of research than
another.
Because different methods serve different purposes, a researcher often will use several in
sequence, so the results from one method can be used by another. For example, in investigating the potential for a new frozen dessert product, a researcher may begin by consulting secondary sources, such as Census statistics or industry trade association statistics, or by studying the performance of similar products that have been launched into the same market. Then
qualitative research would be used to gain insights into the benefits sought by customers and
into sources of dissatisfaction with the existing products. These tentative insights could be confirmed by telephone survey interviews of a representative sample of potential buyers. Finally,
a controlled store experiment might be used to test the appeal of different packages. Data collection methods also vary depending on the managerial style and the culture of the organization. Marketing Research in Action 4-4 describes a method the Japanese use to collect data.
TABLE 4-2
Relationship between Data Collection Method
and Category of Research
Category of Research
Data Collection Method
Exploratory
Descriptive
Causal
Secondary sources
Information system
Databanks of other organizations
Syndicated services
a
a
a
b
b
b
b
a
b
b
a
b
b
a
Primary sources
Qualitative research
Surveys
Experiments
78
a ⫽ Very appropriate method.
b ⫽ Somewhat appropriate method.
RESEARCH APPROACH
MARKETING RESEARCH IN ACTION 4-4
Data Collection—The Japanese Way
Cultural and individual preferences play a major role in determining the research technique and the
method of data collection adopted for a given research project. U.S. managers, in general, prefer gathering large quantities of data through surveys, which provides numbers that can be manipulated statistically. In contrast, managers in Japan prefer the “soft” approach. For example, when Sony conducted a
market survey to determine consumers’ preference for a lightweight portable cassette player, results
showed that consumers would not buy a tape player that did not have the recording function. Regardless of these results, Sony’s chairman Akio Morita went ahead with his plans for introducing the Walkman, and the rest is history.
Sony’s disdain for surveys and other scientific research tools that U.S. managers believe in is shared
by other large Japanese consumer goods manufacturers such as Matsushita and Toyota. Of course, Japanese corporations do want accurate and useful information about their markets; they just go about it differently. Japanese-style market research relies heavily on two kinds of information: “soft data” obtained
from visits to dealers and other channel members; and “hard data” about shipments, inventory levels, and
retail sales. Japanese managers believe that these data better reflect the behavior and intentions of fleshand-blood consumers.
For an article on Japanese data collection methods go to http://hbsp.harvard.edu/product/articles
SOURCE: Adapted from K. Johnny Johansson and Ikujiro Nonaka, “Market Research the Japanese Way,” Harvard Business Review, May–June 1987, p. 16.
Choosing a Research Approach for the HMO Study
Seldom is a data collection method perfectly suited to a research objective. A successful choice
is one that has the greatest number of strengths and the fewest weaknesses relative to the alternatives. Often this is achieved by combining several methods to take advantage of their best features and minimize their limitations. This was what Herb Ellis had to do to get the amount of
information required by the research objectives and still remain within the budget limit (see
Chapter 3 for a discussion of HMOs).
From the beginning it was clear that the overall research approach would involve preliminary qualitative research followed by a survey, to expose the concept of a health maintenance
organization to a large representative sample and test the specific hypotheses. Ellis proposed to
conduct two focus groups to establish the vocabulary used by the target respondents and the
attributes they used to evaluate a health plan, as well as explore their knowledge and expectations of health plans and their reasons for past or prospective changes. The problem was deciding the kind of survey to conduct.
The principal survey options were mail questionnaires and personal or telephone interviews.
Each, however, had a serious drawback. Personal interviews using trained interviewers were
simply too costly and would have been feasible only with a sample that was too small to identify adequately the differences among the three segments. Telephone interviews would have been
difficult to conduct, both because of the length of the questionnaire and the evident need for
multiple-category questions, which are awkward to communicate orally. The questionnaire could
have been administered by mail, but experience suggested that the response rates would be low
unless substantial incentives and follow-ups were used.
The solution was a self-administered questionnaire, with door-to-door delivery and pickup
by untrained survey assistants. The advantage of the telephone in reaching large samples economically was utilized both to establish contact and then to get agreement to participate. During the initial phone call, arrangements were made to deliver and pick up the questionnaire.
Before the pickup, a reminder phone call was made to ensure that the questionnaire had been
completed. In some instances the respondent was given a stamped, addressed envelope so the
questionnaire could be returned by mail.
The research approach was successful in achieving a high response rate at a low cost per
completed interview. The key to success was in matching the approach to the objectives of the
79
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
study and the characteristics of the population, notably, the presence of an up-to-date listing, the
limited geographic area to be covered, and the participants’ inherent interest in the subject of the
survey.
RESEARCH TACTICS AND IMPLEMENTATION
Once the research approach has been chosen, research tactics and implementation follow: the
specifics of the measurements, the plan for choosing the sample, and the methods of analyses must
be developed.
Measurement
The first step is to translate the research objective into information requirements and then into questions that can be answered by anticipated respondents. For example, one of the objectives in the
HMO study is to estimate probable demand for the proposed HMO. This means that information
will be needed on (1) the respondents’ overall evaluation of the proposed HMO, (2) their preference for the proposed HMO relative to their present health plan, and (3) their likelihood of adopting the new plan if it becomes available. As we will see in Chapters 10, 11, and 12, there are many
ways to ask questions to obtain this kind of attitudinal information.
Once the individual questions have been decided, the measuring instrument has to be developed. Usually this instrument is a questionnaire, but it also may be a plan for observing behavior or recording data. The researcher designing an effective questionnaire must be concerned with
how questions on sensitive topics such as income can be asked, what the order of the questions
should be, and how misinterpretations can be avoided.
Sampling Plan
Most marketing research studies are limited to a sample or subgroup of the total population relevant to the research question, rather than a census of the entire group. The sampling plan describes
how the subgroup is to be selected. One approach is to use probability sampling, in which all population members have a known probability of being in the sample. This choice is indicated whenever it is important to be able to show how representative the sample is of the population. Other
critical decisions at this stage are the size of the sample, as this has direct implications for the project budget, and the means of minimizing the effect on the results of sample members who cannot
be reached or who refuse to cooperate.
Anticipating the Analysis
80
When one is bogged down in the details of tactical research problems, it is easy to lose sight of
the research objectives. Before actual data collection begins, the researcher must be alert to the
possibility that the data will be inadequate for testing the hypotheses, or will be interesting but
incapable of supporting action recommendations. Once the data have been collected, it is too late
to lament, “Why didn’t we collect data on that variable?” or “Why didn’t we foresee there
wouldn’t be enough respondents to test that hypothesis?”
With these concerns in mind, the researcher should plan in advance how each of the data items
is to be analyzed. One useful device is to generate fictional (dummy) data from the questions in the
measurement instrument. These dummy data can be analyzed to ensure that the results address the
objectives. For example, a great deal of preliminary data analysis consists of cross-tabulating one
question with a second question. Each of the anticipated tables should be reviewed in terms of its
relevance to the research question. Any shortcomings identified now will help guide the changes
to the questionnaire before it is sent into the field.
RESEARCH TACTICS
AND IMPLEMENTATION
Analysis of Value versus Cost and Time Involved
At this stage of the design, most of the cost has yet to be expended, but the research is now completely specified and a reliable cost estimate should be available. Thus, a more detailed cost–benefit
analysis should be possible to determine if the research should be conducted as designed or if it
should be conducted at all.
One component of cost to be considered is the time involved. A research study can take six
months or more. It may be that such a time period will delay a decision, thus creating the risk that
a set of attractive conditions will be missed. For example, if the research designed to test a new
product takes too long, a competitor may preempt the market with its own version of the product.
The analysis can conclude that either the research design is cost effective and should proceed or
that it is not and should be terminated. Usually, instead of termination, consideration will be given
to a revised research design that will be less costly. Perhaps a smaller sample could be used, or a
laboratory experiment substituted for a field experiment. Throughout the whole research process,
new information is uncovered that may make it useful to alter the purpose, the research question,
the research approach, or some aspect of tactics. Indeed, it is much more accurate to think of the
research process as a series of iterations and reconsiderations rather than an ordered sequence of
well-defined steps. Marketing Research in Action 4-5 illustrates how firms are going after more budgets for marketing research.
Errors in Research Design
The usefulness of a research project depends on the overall quality of the research design and
on the data collected and analyzed based on the design. Several potential sources of error can
affect the quality of a research process. The errors can influence the various stages of the research
process and result in inaccurate or useless research findings. Errors specific to each stage of the
research process are mentioned in subsequent chapters. In Figure 4-2, we present an overview
of the various types of errors that can affect a research design, with a brief description of each
type. For a detailed description of the different types of errors, refer to the Appendix at the end
of this chapter.
MARKETING RESEARCH IN ACTION 4-5
It Pays to be Productive
A 2007 survey by Aberdeen, a research and market intelligence company, revealed that Best-in-Class companies (those performing in the top 20 percent of all survey respondents) have instituted marketing performance
strategies to showcase marketing’s value to the organization and improve management’s view to marketing
decisions. Moreover, 33 percent of all companies surveyed recognize “proof of results for justification of next
year’s marketing budget” as a key driver for the adoption of new performance measurement capabilities.
Best-in-Class organizations showed significant performance improvements in comparison to Industry Average and Laggard companies in several areas over the previous twelve months. Specifically, the
Best-in-Class companies achieved:
■ A 9% mean increase in Return on Marketing Investment (ROMI) in the past 12 months
■ A 12% mean increase of new leads into the sales pipeline in the past 12 months
■ A 11% mean increase of customer cross-sell/up-sell opportunities in the past 12 months
■ A 10% mean increase in customer retention rate in the past 12 months
The key to success for marketing organizations rests in improving marketing performance, and identifying the drivers and detractors for performance. Further, with marketing performance becoming a key indicator for budget allocation, companies are now involved in identifying novel methods to make marketing
activity more effective, efficient, and profitable.
SOURCE: “Best-in-Class Companies Improve Marketing Performance Through Measurement Capabilities,” Market Wire, BNET Publications, November 2007.
81
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
TOTAL Error
Sampling
Error
(Chapter 14)
Design
Errors
•
•
•
•
•
•
Selection error
Population
specification
error
Sampling frame
error
Nonsampling
Error
Administering
Errors
•
•
•
Questioning
error
Recording error
Interference
error
Response
Errors
•
Data error
Intentional
Unintentional
Nonresponse
Errors
Failure to
contact all
members
• Incomplete
responses
•
Surrogate
information
error
Measurement
error
Experimental
error
FIGURE 4-2
Errors in research design.
BUDGETING AND SCHEDULING THE RESEARCH
PROJECT
82
Decisions regarding the allocation of resources—time, money, and human resources—form an
important part of the planning for a research project. In any organization there are constraints
and limitations on the availability and use of resources. Given these constraints, budgeting and
scheduling activities ensure that the resources are used effectively and efficiently. The cost–benefit
analysis that precedes the research design phase gives management a preliminary idea of the
value of a marketing research project. By comparing the expected research information with its
anticipated costs—both time and money—management can decide whether a particular project
is worth conducting. After deciding that the benefits exceed the costs, we enter into the blueprint stage, or the research design stage.
Two common approaches to budgeting for a marketing research project are estimating the
dollar costs associated with each research activity or determining the activities to be performed,
in hours, and then applying standard cost estimates to these hours. The former approach typically is used when a marketing research project is relatively unusual or expensive. The latter
approach is used for routine marketing research projects or when the researcher has considerable knowledge of research activity costs.
Since certain research activities (most notably data analysis) cannot be initiated before other
activities (data collection) are completed, research activities must be closely coordinated for a
research project to be completed on time and within budget. Scheduling makes certain that
appropriate personnel and resources are available to carry out the necessary research activities,
so that the entire research process is completed as economically and efficiently as possible. One
RESEARCH PROPOSAL
ACTIVITY
0 1 3 5 78 9
MEETING HC MKTG. PERSONNEL
DESIGNING QUESTIONNAIRE
OBTAINING HC APPROVAL
PRETESTING QUESTIONNAIRE
REDESIGNING QUESTIONNAIRE
MAKING MULTIPLE COPIES
DEVELOPING CODING MECHANISM
SELECTING SAMPLING FRAME
DATA COLLECTION
DATA ENTRY
DATA ERROR CHECKING
DATA ANALYSIS
REPORT WRITING/TABLES/GRAPHS
OBTAINING FEEDBACK FROM HC
ADDITIONAL ANALYSIS
MAKING CHANGES IN REPORT
SUBMITTING FINAL REPORT
MAKING PRESENTATIONS TO HC
March 23, 2003
TIME PERIOD (Days)
28
33 35 40
RESPONSIBILITY
48 53 56 60 6364
IMC
IMC
IMC
IMC
IMC
IMC
IMC
IMC/SIR
SIR
SIR
IMC
IMC
IMC
IMC
IMC
IMC
IMC
IMC
April 24, 1999
FIGURE 4-3
GANTT chart for the 2003 Cingular Wireless.
approach to scheduling is an activity flowchart, a schematic representation or diagram that
sequences the required research activities.
Scheduling helps marketing researchers answer a vital question: Who is responsible for accomplishing what research activity within what time period? This is a critical question for any marketing research project, for it not only allocates a person to a task but also provides a time frame within
which the task is to be accomplished. Essentially, it identifies the personnel accountable for a particular task.
Several creative managerial techniques can be used for scheduling a research project. The
most often used of these techniques are (1) the critical path method, (2) the program evaluation and review technique, (3) GANTT charts, and (4) graphical evaluation and review
techniques.
The critical path method (CPM) is a network approach that involves dividing the marketing research project into multiple components and estimating the time required to complete each
component/activity. The program evaluation and review technique (PERT) is a probabilitybased scheduling approach that recognizes and measures the uncertainty of project completion
times. GANTT charts are a form of activity flowchart that provide a schematic representation
incorporating the activity, time, and personnel requirements for a given research project. An illustration of the use of a GANTT chart for a marketing research project is given in Figure 4-3.
Graphical evaluation and review techniques (GERT) are essentially a second-generation PERT
approach to scheduling, in which both the completion probabilities and activity costs to be built
into a network representation are considered.
RESEARCH PROPOSAL
A research proposal describes a plan for conducting and controlling a research project. While it
has an important function as a summary of the major decisions in the research process, it is useful for a number of other reasons as well. Administratively, it is the basis for a written agreement
or contract between the manager and researcher, as well as a record of what was agreed upon.
As such, it provides a vehicle for reviewing important decisions. This helps ensure that all parties are still in agreement on the scope and purpose of the research, and reduces later
misunderstandings. Frequently, proposals are used to make a choice among competing suppliers
83
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
and to influence positively the decision to fund the proposed study. For these latter purposes, a
proposal should be viewed as a persuasive device that demonstrates the researcher’s grasp of the
problem and ability to conduct the research, and also highlights the benefits of the study.
Like other communications, the structure and coverage of a proposal must be tailored to the
situation. However, the following content outline has been used widely, as it ensures that likely
questions will be anticipated.
Basic Contents of a Proposal
Executive summary: A brief overview of the contents of the proposal. This may be the only
part some people read, so it should be sufficient to give them a basic understanding of
the proposal.
Research purpose and objective: A description of the management problem, defining the information to be obtained in terms of research questions to be answered. This information
must be related explicitly to the management problem.
Research design: Presents the important features of the research methods to be used, with justification of the strengths and limitations of the chosen method relative to the alternatives. All aspects of the research that might be elements of a contract should be discussed,
such as sample size, quality control procedure, data collection method, and statistical
analysis. Details of questionnaire format, sample selection procedures, and so forth,
should be confined to an appendix.
Time and cost estimates: All negotiated aspects, including total fees, payments, provisions,
treatment of contingencies such as the clients’ decision to expand or cancel the study,
and the schedule for submission of interim, draft, and final reports.
Appendixes: Any technical matters of interest to a small minority of readers should be put at
the back end of the proposal.
An example of a research proposal is given in Marketing Research in Action 4-6.
MARKETING RESEARCH IN ACTION 4-6
Research Purpose
A Research Proposal to Crystal Bank
The purpose of the study is to analyze various issues pertaining to the credit needs
of consumers living in certain areas of Austin that have branches of Crystal Bank.
Crystal Bank is focused on improving customer satisfaction and wishes to undertake a study to continue doing so. Specifically, the research objectives are to
■ Gauge consumer sentiment on the services provided by various financial
institutions;
■ Identify areas of improvement among the services provided by the financial
institutions;
■ Determine the type of loans most requested for/needed by the consumers;
■ Identify the important and attractive attributes of the different types of loans;
■ Determine the types of lending institutions most popular with the consumers;
■ Identify the important characteristics that make a lending institution attractive
to the consumers; and
■ Provide a demographic profile of the consumers.
Research Design
The survey research will be conducted by mailing a questionnaire to a total of 300
consumers living in the areas served by each of the 15 Austin area branches of
Crystal Bank. The questions cover a wide range of issues including
■ Identifying the types of accounts operated by the consumer, whether he or she
is happy with the services;
84
RESEARCH PROPOSAL
MARKETING RESEARCH IN ACTION 4-6 (continued)
■ Obtaining likelihood of the respondent applying for a specific type of loan in
the next 12 months;
■ Determining the characteristics of the loan application and loan repayment that
are important to the respondent;
■ Identifying characteristics of the lending institutions which influence the
respondent’s choice; and
■ Finally, it will seek demographic and lifestyle information about the respondents.
Sample Research Questions
Do you currently deal with a financial institution?
Yes
No
If Yes, what type of financial institutions do you deal with? (Check all that apply)
Savings and Loan
Credit Union
Bank
Other (Please specify
)
Have you applied for any loan in the past five years?
Yes
No
If Yes, where did you apply for the loan? (Check all that apply)
Savings and Loan
Mortgage Company
Other (Please specify
Credit Union
Bank
)
What type of loans did you apply for? (Check all that apply)
Home purchase
Home repair/remodeling
Major appliance purchase
Other (Please specify
Automobile
Educational
Personal
)
Rank the following factors in decreasing order of importance, which will influence
your choice of lending institution in applying for any loan. (1—Most important;
2—Second most important, and so on)
Institution reputation
Location of the institution
Present relationship with the institution
Level of service provided
Other (Please specify
)
Sample Selection
The sampling procedure employed for the project will be stratified sampling, a
technique widely used in such surveys. Stratified sampling ensures representation
of consumers living near each of the 15 locations of Crystal Bank. The quality of
the mailing list will determine the successful implementation of stratified sampling.
Data Collection
Telephone surveys will be used to gather information.
Statistical Analysis
Based on the type of scaled response, appropriate statistical methods will be used
and inferences made. However, given the type of scales which will be used, a substantial part of the analysis will focus on frequencies, relative frequencies, and
cross-tabulations.
Project Report
A written report will be submitted, and an oral presentation of the findings can
be made by our firm.
Project Cost and Schedule
The project will be completed within a period of 12 weeks from the time the information requirements are provided by Crystal Bank to our firm. However, this is
dependent on the assumption that the data collection process will be completed
within a four-week period. The cost of the project is expected to be $10,000.00.
85
CHAPTER 4 RESEARCH DESIGN
AND IMPLEMENTATION
DESIGNING INTERNATIONAL MARKETING
RESEARCH
As we explained earlier in Chapters 1, 2, and 3, international marketing research is conducted
to aid in marketing decisions in more than one country. Designing a research process for international marketing decision making is considerably more complex than designing it for a single country. Conducting research in different countries implies much greater attention to issues
such as:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Understanding the nature and type of information sought
Defining the relevant unit of analysis
Formulating problems, variable specifications, and categories
Identifying and selecting sources of information
Availability and comparability of data
Achieving equivalence of samples and measures across countries and cultures
Identifying the degree of centralization of the research
Coordinating research across countries
Finding errors in the research design
Learning the cost of conducting research in multiple countries
Marketing Research in Action 4-7 gives a brief synopsis of the pitfalls a researcher can encounter
while conducting international marketing research.
MARKETING RESEARCH IN ACTION 4-7
A Practitioner’s View of the Key Pitfalls
in Conducting International Research
The key pitfalls to avoid when conducting an international marketing research project are
1. Selecting a domestic research company to do your international research: Only a handful of domestic
research companies are both dedicated to and expert in international research. It is important that
international projects be coordinated by a team whose sensitivity and knowledge of foreign markets
will ensure a well-executed study. Emphasis should be placed on selecting a research company with
a solid reputation and extensive experience in the design, coordination, and analysis of global
research.
2. Rigidly standardizing methodologies across countries: Attempting to be consistent with a methodological approach across countries is desirable but, among other things, two key questions need
to be asked in order to determine whether a particular methodology will yield the best results:
(a) Does the culture lend itself to that methodology? For example, relationships in Latin America are based on personal contact. Hence, when conducting business-to-business surveys, personal interviews, though expensive, are more efficient than telephone interviews. (b) Does the
local infrastructure hinder the use of that methodology? For example, telephone surveys are very
common in the United States, but in Russia the telephone system is notoriously inefficient. For
example, the Moscow office for The Economist conducted an informal study to determine how
ineffective the phone system is. The office kept a log of international calls made in a 30-day
period. A total of 786 calls were attempted, of which 754 resulted in no connection, 6 calls were
cut off halfway through, and 2 were wrong numbers. Also, the cost of using this inefficient system is exorbitant. To install the phone costs $2,865; one year’s service costs $485, and a 1-minute
call from Moscow to London costs $3.30.1
86
ISSUES
IN INTERNATIONAL
RESEARCH DESIGN
MARKETING RESEARCH IN ACTION 4-7 (continued)
3. Interviewing in English around the world: When conducting business-to-business research, even if the executives in the foreign country speak English, interviewing in English might result in inaccurate
responses. Are the subjects comprehending the questions accurately and fully, or are there nuances to
the question that are not being understood? Are their answers to open-ended questions without detail
and richness due to their apprehension about responding in a non-native language? Moreover, has their
attention been diverted to a consideration of accents (theirs and/or the interviewer’s) rather than the
research questions at hand? Hence, even though translating the questionnaire may be costly and time
consuming, it results in more accurate responses.
4. Setting inappropriate sampling requirements: Several country-specific variables influence the selection of
appropriate sampling procedures in multicountry marketing research. For example, although random
sampling is statistically the most reliable technique to use, it may be impractical in a given foreign
market. Reasons may include the fact that in many of the less developed countries the literacy rate
is very low. Hence, when sampling for surveys that require the respondent to be literate, random
sampling might not work.
5. Lack of consideration given to language: Translations into the appropriate local languages need to be
checked carefully. When possible, a quality control procedure of “back-translation” should be followed. The prime consideration is to ensure translation of the questionnaire so that there is equivalent meaning and relevance in all the countries where the project is being conducted.
6. Lack of systematic international communication procedures: One of the biggest problems of international
research is communicating clearly with the local research companies. Do they understand the objectives of the study? Do they understand the sampling criteria? And do they understand what is
expected from them? All too often, assumptions are made concerning the above issues that lead to
major problems in the study’s execution.
7. Misinterpreting multicountry data across countries: Analysis of the study’s data must focus on the international market from which the data were gathered. Survey comparisons across countries should be
made with the understanding of how the particular countries may differ on many key factors, including local market conditions, the maturity of the market, and the local competitive framework for the
study category.
8. Not understanding international differences in conducting qualitative research: When conducting qualitative
research such as focus groups, group discussions, and in-depth interviews, the researcher must be aware
of the importance of culture in the discussion process. Not all societies encourage frank and open
exchange and disagreement among individuals. Status consciousness may result in situations in which
the opinion of one is reflected by all other participants. Disagreement may be seen as impolite, or certain topics may be taboo.2 Also, in some countries, such as parts of Asia, mixed-sex and -age groups
do not yield good information in a consumer group discussion. Younger people, for example, often
defer to the opinions of older people. If groups cannot be separated by age and sex, one-to-one interviews should be done.
For more information on CASRO go to http://www.casro.org.
SOURCE: Adapted from Daphne Chandler, “8 Common Pitfalls of International Research,” The Council of American Survey Research Organizations Journal,
1992, p. 81.
ISSUES IN INTERNATIONAL RESEARCH DESIGN
Regardless of the basic research design selected (exploratory, descriptive, or causal), researchers
need to be familiar with and experienced in handling several issues or problems unique to the
conduct of marketing research within and across countries and cultural groups. Three issues critical to international research design are (1) determining information requirements, (2) determining the unit of analysis, and (3) achieving equivalence of construct, measurement, sample, and
analysis.3
87
CHAPTER 3 RESEARCH DESIGN
AND IMPLEMENTATION
Determining Information Requirements
In determining the information required for international marketing research, a primary consideration is the specific level and type of decision for which the research is being conducted. In
general, the types of decisions fall into two broad categories—strategic decisions and tactical
decisions. These two types differ significantly in their information requirements and the level in
the organization structure where the decision making is done.
Global strategic decisions are made mostly at corporate headquarters, and they normally concern issues pertaining to foreign market selection, market entry, mode of entry, market expansion
strategies, and decisions related to global standardization versus local adaptation of marketing-mix
strategies. Such decisions involve the entire organization and determine the overall allocation of
company resources across country markets. If a firm is involved in more than one product category, the decisions involve not only country markets but also product markets within countries.
The information required for global strategic decisions is governed by the company’s overall objectives and has implications pertaining to the company’s long-term survival.
Tactical decisions, on the other hand, are concerned with micro-level implementation issues, and
the information required for tactical decision making is obtained mostly from primary data. These
decisions are concerned primarily with marketing-mix strategies in country/product markets—for
example, what type of advertising copy would be effective in a given culture. The decisions are
made at the functional or subsidiary level rather than at the corporate level.
Unit of Analysis
In conducting marketing research in more than one country, another major issue to be sorted out is
at what level the analysis is to be done. Should it be done at (1) the global level, considering all
countries simultaneously (a very complicated and seldom undertaken unit of analysis); (2) the
regional level, considering groups of countries as being relatively homogeneous in terms of macroenvironmental factors (for example, the European Union and the North American Free Trade Agreement countries can be considered regional trading blocs); (3) the country level, where each country
is taken as a separate unit; or (4) similar segments across countries (a recent trend that is gaining
popularity). In this last type of analysis, the researcher targets homogeneous segments having similar tastes and preferences, across countries.
Construct, Measurement, Sample, and Analysis Equivalence
While conducting research across national boundaries, it is important for the researcher to be
able to compare data across countries. Hence, examining the various aspects of data collection
processes and establishing their equivalence becomes a necessity. The following example illustrates the importance of equivalence issues in international marketing research. Initial data from
a study conducted in Europe suggested that the percentage of Belgian women taking baths was
far higher than for any other nationality. However, a closer look at the data revealed that the
time period was not comparable. While women in Belgium were asked if they had taken a bath
in the last seven days, in all other countries the question had been, “Have you had a bath in the
last three days?”4
Construct equivalence deals with how both the researcher and the subjects of the research
see, understand, and code a particular phenomenon. The problem confronting the international
researcher is that, because of sociocultural, economic, and political differences, perspectives may
be neither identical nor equivalent. The international researcher is constantly faced with the selfreference criterion problem and its implications in formulating a research design. Construct
88
SUMMARY
equivalence is concerned with the question: “Are we studying the same phenomenon in countries X and Y?” For example, in the United States, bicycles are used predominantly for recreation;
in the Netherlands and various developing countries, they provide a basic mode of transportation. This implies that the relevant competing product set must be defined differently. In the
United States it will include other recreational products, whereas in the Netherlands it will
include alternative modes of transportation.
Measurement equivalence deals with the methods and procedures the researcher uses to
collect and categorize essential data and information. Construct and measurement equivalence
are highly interrelated. Measurement is the operationalization of the constructs to be used. Measurement equivalence is concerned with the question: “Are the phenomena in countries X and
Y measured the same way?” For example, while Americans measure distance in miles, in most
of the other countries of the world it is measured in kilometers.
Because of sociocultural, economic, and political differences among or between countries, the
international researcher faces two problems not encountered by the domestic researcher: (1) identifying and operationalizing comparable populations, and (2) selecting samples that are simultaneously representative of other populations and comparable across countries. Sampling equivalence
is concerned with the question: “Are the samples used in countries X and Y equivalent?” For example, children in the United States are legitimate respondents, because they exercise substantial influence in the purchase of cereals, toys, desserts, and other items, whereas in Asian cultures it is the
parent who decides most of these issues.
Regarding analysis equivalence, respondents from different countries have a tendency to
choose either extreme scale points or middle values based on their cultural habits. Therefore, it
may not be appropriate to compare the means of different scale items across countries; rather,
some measure of deviation from the norm should be established. In other words, standard deviations may be better measures for comparison purposes.
Apart from these issues, other aspects of the research process, such as identifying sources of
data, availability, and comparability of data from different countries, problems associated with
primary data collection across countries, and so forth, add to the complexity of the international
research process. Also, these issues add to the nonrandom error component of the research
process. These issues will be dealt with in greater detail in subsequent chapters.
Summary
In this chapter the focus has shifted from the manager’s problems and information needs—as expressed in
the research purpose and objectives—to the strategic and tactical decisions that will achieve the objectives
of the research approach. The various research approaches include qualitative research, surveys, observation, and experimentation. Tactical research design decisions include the choice of a research supplier, questionnaire development, the design of the experiment, the sampling plan, and the anticipation of data analysis. Implementation involves a final cost–benefit check, plus data collection, data analysis, and the
development of conclusions and recommendations. Also, issues relevant to the design of international marketing research projects are discussed in this chapter.
An important distinction can be made among exploratory, descriptive, and causal research. Exploratory
research, which tends to involve qualitative approaches such as group interviews, is usually characterized
by ill-defined or nonexistent hypotheses. Descriptive research, which tends to use survey data, is characterized by tentative hypotheses that fall short of specifying causal relationships. Causal research, which
tends to rely on experimentation, involves more specific hypotheses involving causal relationships. Possible sources of errors in research designs are presented, and the concepts of budgeting and scheduling a
research project are discussed in some detail.
The major decisions during the research process are summarized in the research proposal. This step is
essential to ensuring that the manager’s problems have been translated into a research study that will help
obtain relevant, timely, and accurate information—and not cost more than the information is worth.
89
Questions and Problems
1. Is a research design always necessary before a
research study can be conducted?
2. In what ways do exploratory, descriptive, and causal
research designs differ? How will these differences
influence the relative importance of each research
approach at each phase of the marketing program
development process described in Chapter 1?
3. What alternative research approaches should Herb
Ellis consider for the HMO study? What are the strengths and weaknesses of the possible approaches?
4. A manufacturer of hand tools uses industrial
supply houses to reach its major markets. The company is considering a new, automatic inventorycontrol procedure. How would you proceed with
an exploratory study in advance of a larger study
of the dealers’ reactions to this new procedure?
5. What problems can you foresee in a test of the
hypothesis that federal food stamps issued to lowincome individuals are being used to supplement
food budgets rather than replace former spending
on food?
6. The problem of a large Canadian satellite TV company was described in Chapter 3. A number of
hypotheses were offered by management to account
for the poor penetration in several areas comprising
15 percent of the population of the total service area.
If you were the researcher assigned to study this
problem, how would you proceed? Specifically is
the statement of purpose of the research adequate?
What alternative research designs should be considered? Will one design be adequate to test all the
hypotheses?
7. Smith Computers, Inc., a U.S.-based manufacturer of
personal computers, has developed a microcom-
puter using the Pentium microchip technology, but
at a fraction of the cost of its competitors. The company has an in-house marketing research department, and a study has been ordered to assist in
developing the marketing program for this product.
a. Which type of research would be most appropriate for this study?
b. What are the possible errors that could be made
in designing the research project?
c. Scott Peters, the head of the marketing
research department, must prepare a research
proposal. Suggest a content outline that will
ensure that all likely questions will be
addressed.
d. Peters has also been given the task of identifying foreign market opportunities for this product. What critical issues must be considered in
formulating the research design?
8. What possible problems might be encountered by
a domestic research company in conducting an
international research study?
9. a. How is a cost–benefit analysis useful to the
management in deciding whether or not to conduct a marketing research study?
b. What are the two approaches to budgeting for
a market research project?
c. For what situation is each approach most suitable?
10. Sugar Land Creamery is planning to launch a new
flavor of ice cream and wants to get a “snapshot” of
the potential market. The ice cream has a coconutwhite chocolate flavor with mixed-in pistachios and
is aimed at the premium market. What type of
research design is appropriate? Develop the research
purpose, research questions, and hypothesis.
End Notes
1.
2.
3.
“Hung Up,” The Economist, July 20, 1991, p. 50.
R. Michael Czinkota and A. Ronkainen Ilkka, International Marketing, 3rd ed. Orlando, FL: The Dryden
Press, 1993, pp. 550, 551.
V. Kumar, International Marketing Research, Englewood Cliffs, NJ: Prentice-Hall, 2000.
4.
C. Min-Han, Byoung-Woo Lee, and Kong-Kyun
Ro, “The Choice of Survey Mode in Country Image
Studies,” Journal of Business Research, No. 29,
February 1994, pp. 151–152.
Appendix
Errors in Research Design1
The total error in a research study is the difference between the true mean value (within the population)
of the variable being studied and the observed mean value obtained through the research study. This error
has two main components:
1
Figure 4-2 provides a schematic diagram of the various types of errors possible in a research study.
90
Sampling Error
Sampling error is the difference between a measure obtained from a sample representing the population and the true measure that can be obtained
only from the entire population. This error occurs because no sample is a
perfect representation of a given population, unless the sample size equals
the population. This issue will be dealt with in greater detail in Chapters 14
and 15.
Nonsampling Error
Nonsampling error includes all other errors associated with a research
project. There may be several different reasons for these errors, which can
be broadly classified into four groups: (1) design errors, (2) administering
errors, (3) response errors, and (4) nonresponse errors.
Design Errors
Design errors, also called researcher-induced errors, are mainly due to flaws
in the research design. There are several different types of design errors.
Selection Error
Selection error occurs when a sample obtained through a nonprobability
sampling method is not representative of the population. For example, if a
mall interviewer interested in shopping habits of the visitors to the mall
avoids interviewing people with children, he or she is inducing a selection
error into the research study.
Population Specification Error
Population specification error occurs when an inappropriate population is
chosen from which to obtain data for the research study. For example, if the
objective of a research study is to determine what brand of dog food people
buy for their pets, and the research draws a sample from a population that
consists predominantly of cat owners, a population specification error is
induced into the study.
Sampling Frame Error
A sampling frame is a directory of population members from which a sample is selected. A sampling frame error occurs when the sample is drawn
from an inaccurate sampling frame. For example, if a researcher interested
in finding the reasons why some people have personal computers in their
homes selects the sample from a list of subscribers to PC World, he or she
is inducing a sample frame error into the study.
Surrogate Information Error
Surrogate information error is the difference or variation between the information required for a marketing research study and the information being
sought by the researcher. The famous (or rather infamous) New Coke taste
tests are a classic example of surrogate information error. The researchers
in that case were seeking information regarding the taste to New Coke versus Old Coke, but the study should have determined consumer’s attitudes
toward a change in the product and not just their taste preferences.
Measurement Error
Measurement error is the difference or the variation between the information sought by a researcher for a study and the information generated
by a particular measurement procedure employed by the researcher.
Measurement error can occur at any stage of the measurement process,
from the development of an instrument to the data analysis and interpretation stage. For example, if a researcher interested in the individual
income of the respondent words the question as annual household income,
a measurement error is being induced into the research study.
Experimental Error
An experiment is designed to determine the existence of any causal relationship between two variables. Any error caused by the improper design
of the experiment induces an experimental error into the study.
Data Analysis Error
Data analysis error can occur when the data from the questionnaires are
coded, edited, analyzed, or interpreted. For example, incorrect coding of
data or a wrong use of a statistical analysis procedure can induce a data
analysis error into the study.
Administering Errors
Questioning Error
All errors that occur during the administration of a survey instrument to
the respondents are classified as administering errors. They are caused by
mistakes committed by the person administering the questionnaire. They
may be caused by three major factors.
This error arises while addressing questions to the respondents. If the interviewer does not word the question exactly as designed by the researcher,
a questioning error is induced.
91
(continued)
Recording Error
This error arises from improperly recording the respondent’s answers. If
the interviewer misinterprets the response or hears it inaccurately, this
induces a recording error into the study.
Interference Error
This error occurs when an interviewer interferes with or fails to follow the
exact procedure while collecting data. For example, if the interviewer fabricates the responses to a survey, it induces an interference error.
Response Errors
Response errors, also called data errors, occur when the respondent—
intentionally or unintentionally—provides inaccurate answers to the survey questions. This might be due to the respondent’s failing to comprehend
the question or it may be due to fatigue, boredom, or misinterpretation of
the question. Response errors can occur when a respondent who is unwilling or embarrassed to answer a sensitive question provides an inaccurate
or false response.
Nonresponse Errors
Nonresponse errors occur if (1) some members of a sample were not contacted, and hence their responses were not included in the study; or (2) some
of the members contacted provide an incomplete or no response to the survey instrument. The primary reasons for this error occurring include the
unwillingness of respondents to participate in the study and the inability of
the interviewer to contact the respondents.
CASE 4-1
Reynolds Tobacco’s
Slide-Box Cigarettes
There is a tendency for management to think of marketing research as an expense, perhaps a necessary
expense, but an expense nevertheless. With corporate
attention being focused more and more on the bottom
line, one question is often asked: “What is the impact
of current or potential new activities on profit?” This
case focuses on how marketing research can aid in decision making and enhance the bottom line.
To give you some basic understanding of the business situation, the cigarette industry is very mature.
Many big opportunity areas for product differentiation
have already been thoroughly explored and mined by
the major manufacturers. There are different product
attributes, ranging from different flavors (menthol and
nonmenthol), to different “tar” levels (full flavor,
light/low “tar,” ultralow “tar”), to different lengths
(king size to 120 mm), and even different circumferences (superslims to wide circumference). Imagery is
another means of differentiating products, so there are
upscale cigarettes, masculine cigarettes, feminine cigarettes, and so on. Finally, price is another important
dimension differentiating brands. Combining all these
methods of differentiating products results in literally
hundreds of types of cigarettes available to smokers,
ranging from superslim, ultralow-tar 120s in a box to
traditional nonfilter 70-mm cigarettes.
With this in mind, you would think that there are
few new ideas under the sun for tobacco manufacturers.
92
That may be, but ideas that have the potential to draw
business from competition and increase one company’s
share of the market are constantly being examined.
One such new idea was a brand that R.J. Reynolds
Tobacco Co. (RJR) had in test market. This new brand
was developed to capture Marlboro smokers with
imagery that was more sociable and had more appeal
to both male and female adult smokers than Marlboro’s
traditional solitary cowboy image. From a product
standpoint, the new brand’s smoothness was emphasized in contrast to the rich flavor emphasized in
Marlboro advertising. Marlboro’s conventional crushproof box has always been a key feature of the brand.
Rather than offering the new brand in an identical
packaging configuration, Reynolds Tobacco’s R&D
department developed a slide-box, a unique configuration that operated something like a matchbox. This feature was intended as yet another point of difference to
attract Marlboro smokers to the new brand.
Concept product test results of the new brand
(Table 4-3) in a conventional crush-proof box, versus the
slide-box, indicated that the slide-box would enhance
the new brand’s appeal among prospect smokers.
When Reynolds management saw these results,
they said, “If this slide-box is better than the conventional crush-proof box, we shouldn’t just limit it to a
new brand that may or may not make it out of the test
market. To really capitalize on the apparent appeal of
the slide-box, we should put it on our established products.” The direction to put it on established brands is
not as clear as it may sound at first blush. There were
several questions to be resolved: Which brands? Which
styles of which brands? Should the slide-box replace
TABLE 4-3
Concept Product Test Results
Conventional
Crush-Proof Box
Slide-Box
Positive purchase interest
56S
64
Intended frequent use
As usual brand
Occasionally
59S
20
39
68
24
44
Package increased
purchase interest
Increased a lot
Increased a little
51S
27
24
73
53
20
Overall positive taste rating
72S
80
S ⫽ Significant at 80% confidence level or greater.
any current conventional crush-proof box, or should it
be an additional offering? How would a slide-box
launch fit in with other planned programs? How big an
announcement should this be? (All media forms? Pointof-sale advertising only?) Should pack graphics be
redesigned to emphasize the change, or should they
just be translated from existing graphics? How much volume would the slide-box generate and, therefore, how
much new equipment would need to be purchased?
Given the excitement over this proposition, tentative plans were made regarding which brands would
use the slide-box, volume projections were calculated,
and equipment was ordered. The long equipment lead
time provided time to resolve some of the key
questions surrounding the slide-box. The most critical
question was which brand or brands should use the
slide-box. The issue here was one of image compatibility, that is, whether the inherent image of the slide-box
was compatible with the current or desired image of
the brand. Another critical question was what was its
source of appeal. If it was indeed a better package,
what were specific benefits that could be advertised to
prospect smokers?
A good deal of what marketing researchers do is
implement standardized methodologies to answer
recurring marketing questions: How does a product
compare to competition? How effective is this new ad
campaign? How appealing is this premium? The issues
and questions surrounding the slide-box were new, so
the marketing research department developed a custom research design that they felt would generate
required consumer input on the slide-box.
The research was designed with two objectives.
The first objective was to determine which of RJR’s
major established brands has an image consistent with
the image projected by the slide-box; and the second
objective was to determine the degree of interest in this
new slide-box packaging among each of these brands’
prospect smoker groups. In designing the study, they
knew that they needed to analyze the data by each
brand’s prospect group. So they designed the sample
in two phases. First, a large random sample of adult
smokers was drawn. Then additional interviews were
conducted to give them large enough bases of specific
types of smokers. As a result, the sample for the study
consisted of 767 adult smokers, 600 of whom were
selected randomly, and an additional 167 interviews to
fill out specific prospect group quotas.
Given the sample size and the fact that RJR wanted
representation across the country, the marketing research
department conducted personal interviews in 20 geographically dispersed markets during June 1991. The
biggest challenge in designing this research was to evaluate the image of the packaging itself without allowing
any influence of the image of any brand. Hence, they
showed smokers a sample of the packaging without any
graphics. Further, since they wouldn’t be able to evaluate people’s responses to the slide-box packaging in a
vacuum, they also showed them prototypes of current
packaging, that is, conventional crush-proof box and soft
pack, also without any graphics. This provided benchmarks to help them interpret the results.
They agreed that there are two aspects of packaging
on which the packs should be evaluated: benefits of the
packaging itself (such as protecting the contents, being
easy to open or close, etc.), and the image of the types of
smokers who would use such packaging (such as whitecollar versus blue-collar). They showed respondents prototypes of each of the three types of packaging attributes
and user imagery. Importantly, each smoker interviewed
was given a fresh prototype and allowed to open it. After
they had obtained this information, they then had each
93
TABLE 4-4
Overall Evaluation of Packs
Slide-Box
Pack
(%)
Conventional
Crush-Proof Box
(%)
Soft Pack
(%)
Favorable
60
70
42
Unfavorable
26
8
26
of the brands in their study rated on these same user
characteristics.
Finally, they asked the smokers in their study
directly how appropriate the three types of packages
were for the brands of cigarettes included in their study.
The order in which all packs and brands were presented
was rotated to avoid any order bias. The results of the
overall evaluation of packs are given in Table 4-4.
The outcomes of the perception studies are illustrated in Figures 4-4, 4-5, and 4-6.
The researchers found that the slide-box had functionality problems as compared to the conventional
crush-proof box. It was seen as having cosmetic advantages, such as being attractive, modern, being in its
own case, and having an edge over the conventional
crush-proof box. And, once opened, hard to keep
closed. Thus, the cigarettes were more apt to fall out.
The slide-box was also seen as less convenient to use.
And importantly, the slide-box was perceived to be
more gimmicky than the conventional crush-proof box.
It is important to note that although these problems
were pervasive, in that all target groups felt this way,
it was conceivable that these pack problems might go
away as smokers became more experienced with using
this configuration.
It was also important that the user image of the
slide-box was, with one exception, not felt by respondents to be compatible with the user imagery of the
brands for which the pack was being considered. The
analysis found that only one of the established brand’s
prospect group was most likely to find the slide-box as
appropriate for the brand as the conventional crush-proof
box and to believe that brand’s smokers and slide-box
users share many similar characteristics. Both groups
were pictured as white-collar, upper-class women seeking changes and new experiences. The only inconsistency
was that respondents perceived the slide-box user as
European. Their image of the slide-box user was more
consistent with their image of their current brand than
their image of the conventional crush-proof-box user.
The results of the appropriateness of packs of
cigarette brands are presented in Table 4-5.
The new brand was not included in this national
study, as it was only in test market at this time.
However, the new brand’s prospect group did not rate
the slide-box as highly overall as they rated the conventional crush-proof box. Figure 4-7 provides the differences in perception of slide-box-pack users among
the new-brand smokers.
Perceptions of slide-box-pack users
Contemporary
Old-fashioned
Seeks change
Resists change
Seeks new experiences
Prefers familiar
Upper-class
Lower-class
White-collar
Blue-collar
Leader
Follower
Balances work & play
Work comes first
57%
6%
56%
7%
51%
6%
36%
5%
34%
6%
31%
10%
23%
8%
FIGURE 4-4
Perceptions of slide-box-pack users.
94
Perceptions of slide-box vs.
conventional crush-proof-box users
Male
30%
11%
27%
Contemporary
Frequent smoker
57%
23%
11%
Balances work
and play
Conventional crush-proof box
Slide-box
22%
23%
FIGURE 4-5
Perceptions of slide-box versus conventional crush-proof-box users.
Perceptions of slide-box vs. soft-pack users
36%
Frequent smoker
11%
34%
Old-fashioned
Prefers familiar
Resists change
8%
29%
6%
28%
7%
34%
Older
Blue collar
12%
22%
6%
Soft pack
Slide-box
FIGURE 4-6
Perceptions of slide-box versus soft-pack users.
TABLE 4-5
Appropriateness of Packs of Cigarette Brands
Percent Saying Extremely/Very Appropriate
Established
Brand A
Established
Brand B
Established
Brand C
Established
Brand D
Most
Often
Brand
Slide-box
32
47
49
39
48
Conventional
crush-proof box
55
64
51
62
73
Soft pack
61
44
39
55
44
95
Perceptions of slide-box-pack users
(Among Dakota smokers)
Seeks change
Resists change
59%
7%
Contemporary
Old-fashioned
Seeks new experiences
Prefers familiar
58%
8%
57%
5%
Upper-class
Lower-class
7%
White-collar
Blue-collar
7%
36%
35%
Balances work & play
Work comes first
23%
8%
FIGURE 4-7
Perceptions of slide-box-pack users (among new brand smokers).
Questions for Discussion
1.
2.
Why are the results of the second study different
from the first?
Based on these findings, what is your recommendation to the company?
CASE 4-2
California Foods Corporation
In early 1990, the international marketing manager at
California Foods Corporation (CFC), Lois Verbrugge,
was considering how to react to the continuing decline
of CFC grape juice sales in the Puerto Rican market. In
1989, the marketing staff in the international division
estimated that sales of CFC grape juice had fallen off by
approximately 30 percent from the previous year. To
determine why this loss of volume had taken place,
extensive consumer research was utilized. But, as of
February, Ms. Verbrugge and her staff had not come up
with any clear-cut remedies for CFC’s problems in the
Puerto Rican market.
Company Background
CFC was a wholly owned subsidiary of the Federation
of Grape Growers’ Associations. The federation purchased the California Foods Corporation in 1956 as
part of a strategy to integrate its business forward into
the processing and distribution of grape products. CFC
continued in 1990 to operate as an agribusiness largely
as it had in 1956. The federation supplied the grapes,
96
Source: This case was adapted with permission from
a case originally written by H. Daniel Murphy and
Mary E. Brownell, “How Research Can Save Your
Company Time and Money!: A Case Study,” The
Council of American Survey Research Organizations
Annual Journal, 1992, pp. 107–112.
and CFC handled all processing and marketing of the
products. CFC’s sales had increased every year since
the takeover by the federation. CFC was generally
considered the foremost leader in the juice industry. It
set the standards for progressive marketing techniques
and new product development for the industry. With
sales reaching a quarter billion dollars in 1989, the
growers and CFC were the largest grape growing, processing, and marketing enterprise in the world.
Originally, CFC had produced only grape-related
products: grape jams, grape jelly, frozen grape concentrate, grape drink, and grape preserves. In recent years,
however, CFC had expanded to include nongrape
products, too. Between 1970 and 1982, CFC introduced
36 new products. In 1990, CFC incorporated a complete
line of fruit juices with a selection of fruit drinks and a
line of fruit-flavored preserves.
CFC’s International Division
CFC distributed an assortment of products to foreign
markets with the majority of sales derived from juices
and fruit drinks. It marketed its products to over
40 countries. Major markets included Puerto Rico,
Mexico, and Japan. CFC products were distributed
by food brokers and distributors to retail stores and
food service institutions. In 1988, the International Division experienced record sales and greater than expected
profitability. Sales slipped slightly during 1989, largely
the result of sales erosion in the Puerto Rican market.
Frozen concentrates represented another competing group that was large and had shown strong growth
in the preceding three years. Again, the miscellaneous/all-others subgroup had shown steady growth.
Perhaps some of CFC grape juice’s loss could be attributable to a shift of sales across generic categories.
The Juice and Drinks Market in Puerto Rico
Most of the juice consumption in Puerto Rico was composed of imported products. Some of the more popular brands competing for market share were CFC,
Seneca, Pueblo, and Grand Union. There was only one
domestic grape juice producer, selling under the name
Richy. Richy had been in business for a few years, but
its impact on the market had been minimal. Table 4-6
outlines the imported volumes of juices and drinks into
Puerto Rico over the last three years.
As the table reveals, grape juice imports (California
Foods’ and others) were declining rather sharply. Still,
the grape juice market was by far the largest juice
market in Puerto Rico.
The “fruit drink” category was quite large too and
was growing, especially the miscellaneous/all-others
subgroup, which included Tang’s imported powdered
grape and orange drinks. Because many Puerto Ricans
equated powdered grape with grape juice, it was possible that at least some of CFC grape juice’s volume loss
could be traceable to these imports, although no hard
evidence existed.
CFC’s Entry into Puerto Rico
CFC’s first experience in Puerto Rico came in the 1950s
when it introduced CFC grape juice. At that point,
grape juice was practically unheard of by the majority
of Puerto Ricans. Despite this, the introduction was a
resounding success and CFC grape juice became the
best-selling juice in Puerto Rico.
Rumor had it that CFC grape juice’s success was
traceable to the Puerto Rican beliefs that grape juice
was good for men’s virility and for women’s hemoglobin during their menstrual cycles. Pseudomedicinal
drinks were concocted by mixing eggs with grape juice.
The resulting mixture was referred to as an “egg
punch.” To take advantage of this seemingly unique
consumer behavior, CFC launched an “egg punch”
campaign in 1985. One television spot showed a young
Puerto Rican man at a disco drinking an egg punch and
subsequently departing with an attractive young
woman. Print advertising featured a mother nursing
her newborn and copy expounding the nutritional
value of grape juice.
TABLE 4-6
Juices and Drinks Imported into Puerto Rico
Thousands of Cases
(not equivalents)
Percent of
Change
1987
1988
1989
1988–1989
Fruit juices
Vegetable juice
Tomato juice
Apple juice
Citrus juice
Nectars
Pineapple juice
Prune juice
Grape juice CFC
Other
20.6
45.5
84.5
203.5
—
22.5
25.8
569.1
40.6
23.4
21.2
109.0
198.7
5.0
22.9
23.3
586.5
37.1
23.9
26.3
105.6
183.4
1.8
29.1
29.5
412.1
26.6
⫹2.1
⫹24.6
⫺3.1
⫺7.7
⫺64.0
⫹27.1
⫹26.6
⫺29.7
⫺28.3
Fruit drinks
RJR
Bordena
Miscellaneous/all othersb
114.1
92.9
260.5
161.0
124.4
296.4
116.3
132.6
356.0
⫺27.8
⫹6.6
⫹20.4
Fruit juice—frozen and concentrated
Citrus Central
184.8
CFC
34.4
Miscellaneous/all others
378.1
236.6
24.4
431.5
219.5
32.5
499.8
⫺7.2
⫹33.2
⫹15.8
a
Includes Orange Burst instant breakfast drink, Wyler’s ades.
Includes Tang powdered grape and orange drinks.
Source: Maritime Reports, Washington, DC. U.S. Government Printing Office, 1990.
b
97
TABLE 4-7
Consumption Results of Sample of Puerto Rican Grape Juice Users
During 1988 and 1989
Previous Users
n ⴝ 45)
(n
Juices
More
Same
Less
Orange
57.7%
28.9%
11.1%
Grape
13.3
37.8
24.5
Pineapple
22.2
26.7
33.3
Current CFC Users
n ⴝ 155)
(n
Don’t Use
More
Same
Less
43.5%
42.2%
11.7%
24.4
38.9
47.4
13.0
0.7
17.8
23.3
29.9
31.1
15.7
2.3%
Don’t Use
2.6%
Grapefruit
15.6
11.1
51.2
22.1
5.2
16.9
45.4
32.5
Fruit drinks
17.7
20.0
35.5
26.8
13.6
29.2
23.3
33.9
Fruit nectar
20.0
35.6
26.7
17.7
13.6
30.5
30.5
25.4
Powdered drinks
31.1
17.8
24.4
26.7
9.1
32.5
34.4
24.0
Grape juice was indeed CFC’s biggest seller in
Puerto Rico. Sales for 1989 were 412,000 cases. Frozen
concentrated grape juice accounted for sales of 32,000
cases during 1989. Other CFC products were Calfood
fruit drink, California instant powdered grape drink,
CFC grape soda, and CFC strawberry soda.
Consumer Research
In order to ascertain the causes of CFC’s rapid decline
in grape juice sales, an “Awareness, Usage, and Attitude Study” was compiled in February 1990 to update
the marketing department’s understanding of Puerto
Rican grape juice consumers. Two hundred personal
interviews were done with people who had used grape
juice during the previous two years. The study was
administered by a Puerto Rican consulting group.
Results are listed in Table 4-7.
The results of the study showed that the demand
for orange juice had increased tremendously since 1988.
Both current and previous study users of CFC grape
juice were drinking much more orange juice by 1990.
In addition, the percentage of respondents who did not
use orange juice was practically nil.
Current users of CFC juice continued to drink
large quantities of grape juice, as the figures reveal. In
fact, 86 percent of all CFC users said that they drank
as much, or more, grape juice in 1989 as they had previously. However, among the previous CFC users,
there were many more who had decreased their consumption of grape juice than had increased it. Therefore, it was implied that they were not switching from
one grape juice brand to another, but drinking more
orange juice instead. Over 57 percent of previous CFC
users drank more orange juice by 1990 than they had
in early 1988.
The main motive for the purchase of grape juice by
mothers in the sample was because their children had
asked for and/or liked it. The study also revealed that
98
Puerto Ricans perceived grape juice to be both tasty
and nutritious. On the negative side, respondents who
were buying less grape juice had a variety of reasons
for not buying it; most notably, very high price and
preference for other juices were mentioned.
It was discovered that previous CFC users replaced
grape juice with three other types of beverages: other
canned juices (pineapple, orange, grapefruit), natural
juices (orange, grapefruit, tamarind, lemon), and carbonated drinks (Pepsi, Coca-Cola, and the like).
Researchers had asked the question “Why aren’t
you using more CFC grape juice?” The most frequent
response indicated that CFC’s price was too high and
that the respondents tried to buy products that were
more economical. Secondary reasons suggested that
they did not like the taste and preferred other flavors
to grape. Table 4-8 summarizes consumers’ reasons for
buying either less or no grape juice in general and of
CFC’s in particular.
CFC had performed a similar consumer study in
1985 to determine grape juice drinkers’ attitudes
toward CFC grape juice. One section of the 1985 questionnaire involved consumers’ opinions of the characteristics of CFC grape juice. Likewise, part of the 1990
survey was devoted to similar questioning. In both
studies, respondents rated CFC grape juice on the basis
of eight criteria, on a scale from 1 to 6. The figures in
Table 4-9 represent average ratings for each of the product characteristics.
Both studies seemed to suggest that CFC grape
juice had been, and still was, well regarded in the
Puerto Rican market. There had not been too much
change in the general opinion that CFC grape juice was
a good-tasting, nutritious, high-quality product. In consumers’ minds even the price had become more reasonable in relation to the generally stormy economic
conditions. So what seems to be the problem with CFC
grape juice in the Puerto Rican market?
TABLE 4-8
Respondents’ Reasons for Not Buying
Grape Juice
High price
Reasons for No
Longer Serving
Grape Juice
Reasons for No
Longer Serving
CFC Grape Juice
22.6%
23.2%
Only use it occasionally
9.7
4.4
29.0
22.2
Harmful to stomach/diet
12.9
10.3
Prefer natural juices
16.1
6.7
Prefer other flavors
Not accustomed to using it
n.a.
8.9
Prefer powdered drinks
n.a.
8.7
Other
9.7
15.6
Total
100.0%
100.0%
TABLE 4-9
Averaged Ratings of CFC Grape Juice
(scale of 1 to 6)
Sweetness
Taste
Economy
Nutrition
Naturalness
Best for children
Best for adults
Quality
n ⫽ 200)
1990 Study (n
n ⫽ 200)
1985 Study (n
4.95
4.96
3.86
5.06
4.91
4.97
4.88
5.13
3.96
4.73
3.47
5.24
5.05
4.92
4.74
5.17
The study data appear to support the notion that
CFC grape juice is held in high esteem in Puerto Rico,
yet a solution to CFC’s sales problem is needed. With
this in mind, Ms. Verbrugge arranged a meeting with
Jeff Hartman, Market Research Manager, to discuss and
review the situation. Ms. Verbrugge wanted to examine
the problem in more detail and was prepared to commit
additional funds for marketing research. Before making
any decision, however, she wanted Mr. Hartman’s
assessment of the situation.
Source: This case was reprinted with permission
from Subash C. Jain, International Marketing
Management, CA: Southwestern Publishing Co., 1993.
Cases for Part I
Nature and Scope of Marketing Research
CASE I-1
Clover Valley Dairy Company
In the fall of 1978, Vince Roth, General Manager of the
Clover Valley Dairy Company, was considering whether
a newly developed multipack carrier for yogurt was ready
for market testing and, if so, how it should be tested.
Since 1930, the Clover Valley Dairy Company had
sold, under the trade name Valleyview, milk, ice cream,
and other milk by-products—such as yogurt, cottage
cheese, butter, skim milk, buttermilk, and cream—in
Camden, New Jersey. The raw milk was obtained from
independent farmers in the vicinity of Camden and was
processed and packaged at the Clover Valley Dairy.
Clover Valley’s sales had grown steadily from 1930
until 1973 to an annual level of $3.75 million. However,
between 1973 and 1977, a series of milk price wars cut
the company’s sales to $3.6 million by 1977. During this
99
time, a number of other independent dairies were
forced to close. At the height of the price wars, milk
prices fell to 75 cents per half-gallon. In the spring of
1977, an investigation of the milk market in Camden
was conducted by the Federal Trade Commission and
by Congress. Since then, prices had risen so that Clover
Valley had a profit for the year to date.
Clover Valley served approximately 130 grocery
store accounts, which were primarily members of a cooperative buying group or belonged to a 10-store chain
that operated in the immediate area. Clover Valley no
longer had any major chain accounts, although in the
past they had sold to several. Because all three of the
major chains operating in the area had developed exclusive supply arrangements with national or regional
dairies, Clover Valley was limited to a 30 percent share
of the Camden area dairy product market.
Although Clover Valley had a permit to sell its
products in Philadelphia, a market six times the size of
Camden, management decided not to enter that market and instead concentrated on strengthening their
dealer relationships. In addition, it was felt that, if a
price war were to ensue, it might extend from Philadelphia into the Camden area.
With the healthier market and profit situation in
early 1978, Clover Valley began to look for ways to
increase sales volume. One area that was attractive
because of apparent rapid growth was yogurt. During
the previous three years, management had felt that this
product could help to reverse Clover Valley’s downward sales trend, if given the correct marketing effort.
However, the financial problems caused by the loss of
the national grocery chains and the price war limited
the firm’s efforts. As a result, Mr. Roth felt that Clover
Valley had suffered a loss of share of yogurt sales in the
stores they served.
Since 1975, Mr. Roth had been experimenting with
Clover Valley’s yogurt packaging with the hope that a
new package would boost sales quickly. All dairies in
Clover Valley’s area packaged yogurt in either 8-oz or
1-lb tubs made of waxed heavy paper. Clover’s 8-oz tub
was about 5 in. high and 212 in. in top diameter, tapering to 134 in. at base.
The first design change to be considered was the
use of either aluminum or plastic lids on the traditional
yogurt tubs. However, these were rejected because the
increased costs did not seem to be justified by such a
modest change. Changing just the lid would not make
their tubs appear different from their competitor’s tubs,
it was felt.
By 1976, Mr. Roth had introduced a completely different package for Clover Valley’s yogurt. The 8-oz tubs
were replaced by 6-oz cups, designed for individual
servings. In addition, the new cups were made of plastic and had aluminum foil lids. The 1-lb tubs were
unchanged. No special promotional effort was undertaken by Clover Valley, but unit sales of the new 6-oz
100
cups were more than triple the unit sales of the old
8-oz tubs (see Exhibit I-1). While the increased sales
volume was welcomed, the new plastic cups increased
unit packaging costs from 7.2 cents to 12.0 cents. This
more than offset the saving of 4 cents because of the
reduction in the amount of yogurt per container. Retail
prices were reduced from 41 cents to 34 cents for the
new 6-oz cup, while the price for the 1-lb tub remained
at 75 cents. The increased sales then increased the total
dollar contribution to fixed costs from yogurt by only
5 percent. (All dairies priced their yogurt to give retailers a 10 percent margin on the retail selling price.
Competitor’s retail prices for their 8-oz tubs remained
at 41 cents.)
Mr. Roth felt that both the change to plastic and
the convenience of the smaller size were responsible for
the increased sales. However, he was disappointed
with the high packaging costs and began to look at ways
of reducing them, without changing the package much
further. He felt another package change would be too
confusing to consumers. Because of the economies of
scale needed to produce plastic containers, costs could
be reduced if more units were produced and sold.
Mr. Roth felt that packaging a number of cups together
would make the 6-oz cups easier to carry home, which
might increase sales, and would certainly reduce packaging costs.
By 1978, work had begun on developing a multipack holder to hold six cups together. A single strip of
aluminized plastic would serve both as holder and as
the top for two rows of three yogurt cups. A single cup
could be readily separated from the others in the pack.
Dairy personnel constructed wooden models of several
different cups for use with the holder and with plasticmolding experts, choosing one that would mold easily
and cheaply. Eventually, some of these carriers were
made to order for testing in the plant and among
Clover Valley employee families.
Several problems soon became apparent. The
holder did not always fasten securely to all six cups in
the multipack. While the holder strip was being put on,
the side walls of the cups were slightly compressed,
causing some cups to crack at the edges. When consumers tried to remove one of the cups, they sometimes
pulled the top from an adjacent cup. The problem was
the strength of the aluminized plastic, which made it
difficult to tear even when perforated.
The multipack was redesigned and again tested in
the plant and by employee families. It appeared that the
new package was performing satisfactorily. Negotiations
with Clover Valley’s carton supplier resulted in an estimated price of 8.5 cents for the first 100,000 units. Thereafter unit costs would drop to 7.5 cents per 6-oz cup.
Mr. Roth decided that the best multipack carrier
presently possible had been designed. His attention
then turned to methods of testing the new packs for
consumer acceptance. Mr. Krieger, his father-in-law
EXHIBIT I-1
Clover Valley Dairy Company: Sales Results
1974
1975
1976
1977
1978
Unit Sales of Yogurt—8-oz Tubs (6-oz after June 1977)
January
February
March
April
May
June
July
August
September
October
November
December
1,740
1,437
1,347
1,203
996
960
853
861
915
978
1,254
1,212
1,485
2,928
3,528
3,531
3,651
3,258
3,888
4,425
4,044
3,546
3,696
3,561
4,731
4,499
6,177
7,899
7,629
6,677
6,081
5,814
12,726a
13,422
15,105
23,601
23,214
22,146
17,916
18,594
20,187
20,676
20,199
18,420
14,424
16,716
16,716
18,657
3,725
3,510
3,344
3,503
3,101
3,537
3,827
3,103
2,871
3,028
2,796
3,086
2,971
3,232
2,866
3,392
2,390
2,094
2,589
2,384
2,895
Unit Sales of Yogurt—1-lb Tubs
January
February
March
April
May
June
July
August
September
October
November
December
a
3,882
4,015
4,061
3,573
3,310
3,252
3,383
3,721
3,415
3,276
3,865
4,110
3,715
3,596
3,670
3,405
3,482
3,376
3,366
3,307
3,275
3,450
4,650
3,908
3,937
3,833
3,285
3,333
3,609
3,366
2,837
2,616
2,729
2,816
3,375
3,386
6-oz tubs.
and president of Clover Valley, sent him the following
letter concerning market testing:
Dear Vince,
Concerning the market test of the new cups and carriers, I
have a few suggestions that may be helpful, although the final
decision is yours. I think we should look for a few outlets
where we are not competing with the other dairies, perhaps
the Naval Base or Bill’s Market. Actually, if we use Bill’s,
then the test could be conducted as follows:
1. Give Bill a special deal on the multipacks for this weekend.
2. In the next two weeks, we’ll only deliver the multipacks and no single cups at all.
3. In the third week we’ll deliver both the packs and the
single cups.
4. During the third weekend we’ll have someone make a
survey at the store to determine its acceptance.
5. Here is how it could be conducted:
a. Station someone at the dairy case.
b. After the shoppers have chosen either single cups
or the multipacks, question them.
c. If they chose the multipacks, ask them why.
d. If they chose the single cups ask them why they
didn’t buy the packs.
e. Thank them for their help and time.
Yours,
CHARLES KRIEGER
(signed)
Questions for Discussion
1. Should the new multipack carrier be tested?
2. If a test is judged necessary, what should be the
criteria for success or failure?
3. How useful is the proposed test in addressing the
management problem? What changes, if any,
would you recommend?
101
5
SECONDARY SOURCES
OF MARKETING DATA
Learning Objectives
■ Define and introduce the various secondary data sources.
■ Discuss the uses, benefits, and limitations of secondary data.
■ Describe the various sources and forms of secondary data.
■ Provide a brief overview of the sources of secondary data used in international marketing
research.
■ Get a feel for the applications of secondary data in domestic and international marketing
research.
The previous chapters identified many data sources. Marketing intelligence can also be gathered
by linking and cross-analyzing existing database information. Many research organizations now
have vast stores of information in their databases that cover customer behavior over time. These
only suggest the vast array of possibilities that can literally submerge the manager and researcher
in numbers. The real problem with this data explosion is not the quantity but the variability of
the sources with respect to quality, availability, cost, timeliness, and relevance to the needs of the
decision maker. In this chapter we begin the task of identifying and assessing the data that
answer the decision maker’s specific questions.
One of the hallmarks of a competent marketing researcher is familiarity with the basic
sources pertaining to the market being studied, coupled with sensitivity to their respective
strengths and weaknesses. This means that time will not be lost in an aimless search for nonexistent data, and neither time nor money will be wasted on a premature decision to go into
the field to obtain the data.
Figure 5-1 shows the principal sources available to a researcher who is responding to a
research question or considering what data to collect in order to anticipate future information
needs. This chapter deals with externally available secondary sources, for which the specification, collection, and recording of the data were done for another purpose by someone other
than the user. We will take a particularly close look at census data, because it is so fundamental
to understanding all aspects of a market economy. This becomes evident when census data are
used to analyze market demand. We will also look at standardized data (Chapter 6), which
are collected especially for a set of information users with a common need. Standardized data
are both purpose-specific and expensive, but still much cheaper than having each user do it
singly. Often, the immediate and unique needs of a decision maker require collecting original,
or primary data, which is the topic of the rest of the book.
SECONDARY DATA
Secondary data are data that were collected by persons or agencies for purposes other than solving the problem at hand. They are one of the least expensive, quickest and easiest means of access
to information. Hence, the first thing a researcher should do is search for secondary data available
102
USES
Primary
data sources
OF
SECONDARY DATA
• Sales/Patronage results (outcomes)
• Marketing activity (inputs)
• Cost information
• Distributor reports and feedback
• Customer feedback
Internal
records
Data
sources
Secondary
data sources
Electronic
• Government
• Trade associations
• Periodicals
• Newspapers
• Books
• Annual reports
• Private studies
Published
data
Printed
External
sources
Standardized
sources of
marketing
data
• Store audits
• Warehouse withdrawal services
• Consumer purchase panels
• Single –source data
• Nielsen's television index
• Starch scores
• Arbitron panel
• Multimedia services
Internet
• Blogs
• Competitors
• Web sites
FIGURE 5-1
Sources of secondary data.
on the topic. The amount of secondary data available is overwhelming, and researchers have to
locate and utilize the data that are relevant to their research. Most search procedures follow a
distinctive pattern, which begins with the most available and least costly sources. Figure 5-1 shows
the various sources of secondary data. The order from top to bottom corresponds roughly to
the order in which the alternative sources should be considered, or to the likelihood of that type
of data being incorporated into the marketing information system. That is, almost all information systems initially are based on routinely collected internal data, and expand through the
inclusion of data from published and standardized sources.
USES OF SECONDARY DATA
Secondary data can be used by researchers in many ways. In this section we look at the various
ways in which it can be used.
1. Secondary data may actually provide enough information to resolve the problem being
investigated. Suppose a marketing researcher needs to know the income of households in a
particular market area; all he or she has to do is to look into the appropriate Census Bureau
report.
103
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
2. Secondary data can be a valuable source of new ideas that can be explored later through primary research.
3. Examining available secondary data is often a prerequisite to collecting primary data. It helps
to define the problem and formulate hypotheses about its solution. It will almost always provide a better understanding of the problem, and its context frequently will suggest solutions
not considered previously.
4. Secondary data is of use in the collection of primary data. Examining the questionnaires,
methodology and techniques employed by other investigators in similar studies may be useful in planning the present one. It may also suggest better methods.
5. Secondary data also helps to define the population, select the sample in primary information
collection, and define the parameters of primary research.
6. Secondary data can also serve as a reference base against which to compare the validity or
accuracy of primary data. It may also be of value in establishing classifications or baselines
that are compatible with past studies so that trends may be more readily analyzed.
BENEFITS OF SECONDARY DATA
The most significant benefits secondary data offer a researcher are savings in cost and time. Secondary data research often involves just spending a few days in the library or on the Internet
extracting the data and reporting them. This should involve very little time, effort, and money
compared to primary research. Even if the data are bought from a standardized data source, it
will turn out to be cheaper than collecting primary data, because the cost of data collection is
shared by all those using the data.
Certain research projects may not be feasible for the firm; in such cases, recourse to secondary data will be the only solution. For example, if a firm needs some information on the
entire population of the United States, it will be neither physically nor financially possible for
the company to obtain it. Historical data are always secondary data. If a firm wants to obtain
information on something that happened in the past, it cannot conduct primary research to
obtain it.
In some cases secondary data can be more accurate than primary data. For example, if a
company wants information on the sales, profits, and so forth, of other companies, it can get
more reliable and accurate information from government-released sources, trade associations, or
general business sources than from the companies themselves.
LIMITATIONS OF SECONDARY DATA
104
Despite the many potential benefits of secondary data, they also have a number of limitations. By
definition, secondary data are data that were collected in the past for purposes other than the current research. Hence, problems of fit sometimes occurs between the data required for current
research and the available data. The available data may have a different unit of measurement from
what is required. For example, consumer income can be measured and reported at the individual,
family, or household level. Even assuming that the data use the same unit of measurement, there
still may be differences in the class definitions. For example, if the problem demands classification
of income of individuals in increments of $10,000 ($0–$10,000, $10,001–$20,000, etc.), it does not
help the researcher if he or she gets data where it is classified in increments of $7,500 ($0–$7,500,
$7,501–$15,000, etc.).
The researchers have no knowledge of how the data were collected, nor do they have any
control over it. Therefore, they do not know anything about its accuracy or its bounds of error.
They must make a number of assumptions before they can use it for actual analysis. It is also
very difficult to evaluate the accuracy of the data already collected, because one can gauge its
accuracy only by assessing such research characteristics as the methodology used or evidence of
INTERNAL SOURCES
OF
SECONDARY DATA
TABLE 5-1
Benefits and Limitations of Secondary Data
Benefits
Limitations
1. Low cost.
1. Collected for some other purpose.
2. Less effort expended.
2. No control over data collection.
3. Less time taken.
3. May not be very accurate.
4. Sometimes more accurate than primary
data.
4. May not be reported in the required form.
5. Some information can be obtained only
from secondary data.
6. May not meet data requirements.
5. May be outdated.
7. A number of assumptions have to be made.
conscientious work. In many cases the secondary data may not be sufficient to meet the data
requirement for the research at hand. In these cases, researchers may have to use primary
research.
Secondary data may be outdated, and hence cannot be used in current research. Data about
attendance at theaters five years ago will probably be irrelevant to determine the type of motion
pictures to be produced next spring, because motion picture preferences continually change.
Another problem frequently faced by researchers using secondary data is one of publication currency. The time from data collection to data publication is often long; hence, the data are outdated even when they are first available. An example is the government census, which takes
three years to be published. The benefits and limitations of secondary data are summarized in
Table 5-1.
Finally, when using secondary data, it is important to go to the original source. By doing so
you can determine who collected the data, why it was collected, how it was collected, and when
it was collected. The answers to these questions are important since they help you better understand the quality of the data. Also, by knowing who collected the data, you have the ability to
contact that source and find out more about the study.
INTERNAL SOURCES OF SECONDARY DATA
Internal Records
A company’s internal records, accounting and control systems, provide the most basic data on
marketing inputs and the resulting outcomes. The principal virtues of these data are ready availability, reasonable accessibility on a continuing basis, and relevance to the organization’s situation.
Data on inputs—marketing effort expended—can range from budgets and schedules of expenditures to salespeople’s call reports describing the number of calls per day, who was visited, problems and applications discussed, and the results of the visit.
Extensive data on outcomes can be obtained from the billing records on shipments maintained
in the accounting system. In many industries the resulting sales reports are the single most
important items of data used by marketing managers, because they can be related (via exception reporting methods) to plans and budgets to determine whether performance is meeting
expectations. Also, they may be compared with costs in order to assess profitability.
New developments in information technology that tie customers more tightly to suppliers are
improving the timeliness and depth of the sales information available to managers. For example,
American Hospital Supply has supplied hospitals with computers so that hospital order entries
go directly to the sales department, where they are stored in a computer and can be immediately
accessed and analyzed for trends and transaction details. Salespeople at Wrangler Womenswear
105
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
can connect their portable computers to the corporate computer to send and retrieve messages,
enter orders, and receive up-to-date sales information.
Using Internal Data Effectively
Many diagnostic studies potentially can be undertaken with various combinations of internal
and external data, to address such questions as the following:
■ What is the effect of marketing inputs (number of sales calls or types of distribution channels)
on outcomes such as profitability and unit sales within regions or sales territories?
■ Is our sales performance within key market segments or types of retailers improving or deteriorating?
■ Are current sales and marketing expenditures above or below the levels set in the annual
budget and sales plan?
Such insightful analyses, however, often are thwarted because of limitations in the accounting
system and distortions in the data.
The first problem is that accounting systems are designed to satisfy many different information needs. As a result, the reporting formats frequently are rigid and inappropriate for marketing decisions. Often the accounting data are too highly aggregated into summary results and
are not available for key managerial units, such as geographic areas, customer types, or product
types. Efforts to break down sales and profitability data by different units may involve special,
time-consuming studies. It is also possible that production, sales, and profit figures are each
measured in slightly different time frames, which are all at variance with external data such as
bimonthly store audit data.
A second problem is the quality of the data found in the internal records. On the input
side, the reports of salespeople’s call activities may be exaggerated if they are being evaluated on this basis. Indeed, the well-known optimism of salespeople may unconsciously pervade all the data from this source. Accounting data are not exempt from such problems. The
usual interpretation of a sales invoice is compromised if liberal return privileges are permitted or if the product is purchased at one location but delivered to or used in another. In general, whenever there is a long distribution channel, with several places where inventories can
be accumulated, the data on orders received or invoices billed may not correspond to actual
sales activity.
Customer Feedback
106
Increasingly, companies are augmenting their internal records with systematic compilations of
product returns, service records, and customer correspondence, in a manner that permits easy
retrieval. Responding to the voice of the customer has become critical in order to maintain or
increase market share in today’s competitive environment.1 Complaint letters are being used as
sources of data on product quality and service problems. One reason is the insight they can
provide into the problems of small groups with unusual requirements, reactions, or problems.
For example, a premarket skin abrasion test of a new talc-based bath powder uncovered no problems, but the complaint letters that poured in shortly after the reformulated product was introduced revealed serious problems among a small group with sensitive skin.
Complaint letters, however, present an incomplete and distorted picture. People who write
such letters are not typical clients or customers. They are most likely to be highly educated, articulate, and fussy, with more than average amounts of free time. A letter of complaint is actually
a rather infrequently used method of resolving dissatisfaction; instead, people are more likely to
switch brands, shop in a different store, or complain to their friends. Manufacturers are almost
completely cut off from knowledge of customer unhappiness, because most complaints are
voiced to retailers and there is little systematic feedback from retailers to manufacturers.
EXTERNAL SOURCES
OF
SECONDARY DATA
Customer Database
Many companies have started to build customer databases on their own. A customer database
is raw information on the customer that can be sorted and enhanced to produce useful information. Records of frequent customers and their transactions are maintained, and the companies use this data to find out what is common among its customers. This data can also be used
to find out about customers’ product preferences, form of payment, and so on. Holiday Inn
has created a customer database for its Priority Club members in order to track their activities
and transactions with regard to the company.2 These customer databases are now being used
extensively by marketing managers for formulating relationship marketing strategies and to
determine the actual value of a customer (CLV). This is discussed in greater detail in the final
chapter of the book.
EXTERNAL SOURCES OF SECONDARY DATA
Published Data Sources
Published data are by far the most popular source of marketing information. Not only are the
data readily available, often they are sufficient to answer the research question. For example:
■ A marketing manager studying developments in the wine industry will use trade association
data to learn how the total consumption of wine is broken down, by type of customers, geographic area, type of wine, brand name, and distribution client. These data are available annually and sometimes quarterly, so significant trends can be isolated readily.
■ A person starting a new specialty shop will use census data on family characteristics and
income to support a likely location for the shop.
■ Local housing planners rely on census data dealing with the characteristics of housing and
households in their locality to judge the need for new housing construction or housing rehabilitation.
The prospective user of published data also is confronted with the problem of matching a
specific need for information with a bewildering array of secondary data sources of variable and
often indeterminate quality. What is needed first is a flexible search procedure that will ensure
that no pertinent source is overlooked, and second, some general criteria for evaluating quality.
These issues will be dealt with in the next two sections.
Finding Published Sources. The major published sources are the various government publications (federal, state, provincial, and local), periodicals and journals, and publicly available reports
from such private groups as foundations, publishers, trade associations, unions, and companies.
Of all these sources, the most valuable data for the marketing researcher come from government
census information and various registration requirements. The latter encompass births, deaths,
marriages, income tax returns, unemployment records, export declarations, automobile registrations, property tax records, and so on.
How should someone who is unfamiliar with a market or research topic proceed? In general, two basic rules are suggested to guide the search effort: (1) Start with the general and go
to the specific, and (2) make use of all available expertise.3 The four main categories are authorities, general guides and indices, compilations, and directories.
Authorities. Knowledge of pertinent sources—and of their limitations—comes from continued
experience. Thus, the best starting point is someone else who has been doing research on the
same subject. Trade associations and specialized trade publications are particularly useful, for
107
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
they often compile government data and collect additional information from their subscribers
or members.4 If information about a specific geographic area is sought, the local chamber of
commerce is a good place to begin. When the problem or topic is too large or ill defined, there
is no substitute for a well-informed reference librarian.
General Guides and Indices. Within the category of guides and indices there is a hierarchy of
generality. At the top are the “guides to the guides,” such as Constance Winchell’s Guide to Reference Books (Chicago: American Library Association), The Bibliographic Index: A Cumulative Bibliography of Bibliographies (New York: H. W. Wilson Co.), The Cumulative Book Index (New York: H. W.
Wilson Co.), and Guide to Special Issues and Indexes of Periodicals (New York: Special Libraries
Association).
At the next level of reference materials are guides to general business information sources.
Several important bibliographies are the Encyclopedia of Business Information Sources, the Encyclopedia of Geographic Information Sources (Detroit: National Gale Research Co.), and the Statistical
Reference Index (Washington, D.C.: Congressional Information Service). At a third level of generality, business periodical indices (e.g., Psychological Abstracts, Washington, D.C.: American Psychological Association; The Wall Street Journal) contain references to a large number of journals,
periodicals, and newspapers.
For studies of international markets there is the International Bibliography of Marketing and Distribution (Munchen-Pullach: Verlag Dokumentation). Each country has its own reference guides to
domestic periodicals. For example, in Canada there is the (Annual) Statistics Canada Catalogue and
the Canadian Business Index, and in the United Kingdom there is the Annual Abstract of Statistics
(London: H. M. Stationery Office). Most countries have reference guides to state and provincial jurisdictions. For example, Canadiana (Ottawa: National Library of Canada) includes a regular listing
of provincial and municipal government publications.
Marketing researchers often overlook valuable information on trends and conditions in
specific markets, which is produced by firms such as Frost and Sullivan, Predicasts, Euromonitor, Economist Intelligence Unit, Stanford Research Institute, and A. D. Little. Although these
reports may be expensive, they usually are much cheaper than primary research. These reports are
indexed and described in:
■ Findex (Find/SVP) provides fully integrated research, advisory, and business intelligence services in a broad range of industries and disciplines.
■ Research Alert Direct, a new line of Internet-delivered alerting services. Reports deal with
trends in consumer markets, attitudes, and lifestyle.
Compilations. Compilations are intermediate sources, in that they facilitate access to the original
sources. This is particularly desirable with statistical information. The standard work in this area
is the Statistical Abstract of the United States (Washington, D.C.: U.S. Bureau of the Census), which
contains selections from the various censuses as well as data collected by other agencies. For example, data on the number of industrial robots installed worldwide, by country, are compiled by the
U.S. International Trade Commission. General-purpose marketing statistics are published in volumes such as Market Guide, Marketing Economics Guide (New York: Marketing Economics Institute), and the Rand McNally Commercial Atlas and Marketing Guide. Other valuable compilations
are the Sales and Marketing Management annual statistical issues detailing the “Survey of Buying
Power,” which includes industrial incomes, sales of six types of retail stores, market potential
indices for states, countries, and metropolitan areas, and similar statistics for Canada.
108
Directories. Directories are useful for locating people or companies that could provide information. Trade directories supply a wealth of information on individual companies, including
addresses, names of executives, product range, and brand names. Information on parent companies and/or subsidiaries often is provided. The Thomas Register of American Manufacturers
(New York: Thomas Publishing Co.) lists such data on more than 150,000 manufacturing firms.
EXTERNAL SOURCES
OF
SECONDARY DATA
There is now a Thomas Register for Europe, which is the first pan-European buying guide for
manufacturers. Who Owns Whom (North American Edition) lists 6,500 parent companies and
100,000 domestic and foreign subsidiaries and associated companies. Some directories are narrowly focused, such as McKitrick’s Directory of Advertisers, or the Pulp and Paper Directory of
Canada. A number of directories, such as World Who’s Who in Finance and Industry (Detroit:
National Cole Research Co.) and Standard and Poor’s Register of Corporations, Directors and Executives (covering the United States and Canada), provide general biographical information on
individuals.
It is important to realize that only a few of the better-known sources have been described or
mentioned here. Almost all of these guides, compilations, and directories are now available either
online or on CD-ROM. The researcher is always best advised to seek the assistance of a qualified reference librarian whenever a new area or topic is being studied.
Computer-Retrievable Databases
Even with the array of printed bibliographies, directories, and indices, a search can be very timeconsuming. Recent advances in computer technology have resulted in more efficient methods of
cataloging, storing, and retrieving published data. The growth in the number of databases available electronically through computers has been dramatic. Thousands of online databases are available to researchers and analysts working in almost every area of business, science, law, education,
and the social sciences. These databases are easily accessible from personal computers and PDA
devices. Increasingly, the interfaces developed for the user’s communication with the database system is designed to be user friendly. As a result, use of these electronic information sources has
expanded rapidly to facilitate almost any search for information and is no longer limited to computer specialists.
The large number of databases can be overwhelming. Databases can be classified by type of
information contained or by the method of storage and retrieval. Figure 5-2 gives a comprehensive view of the classification.
Computer-retrievable
databases
Based on the method of
storage and retrieval
of information
Online
databases
Internet
Direct from
vendors
CD-ROM
databases
Direct from
producer
FIGURE 5-2
Classifying computer-retrievable databases.
Based on the type
of information
Source
Reference
Indirect
through
networks
109
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
Classification of Databases Based on Content of Information
Reference Databases. Reference databases refer users to articles and news contained in other
sources. They provide online indices and abstracts and are therefore referred to as bibliographic databases. Use of these databases is a quick and efficient method for researching a
subject before obtaining a large amount of detailed information. Reference databases provide
three distinct search features:
1. They are up-to-date summaries or references to a wide assortment of articles appearing in
thousands of business magazines, trade journals, government reports, and newspapers
throughout the world.
2. The information is accessed by using natural-language key words, rather than author or title.
For example, the word “steel” will cause the computer to retrieve all abstracts that contain
that word.
3. Key words can be combined in a search for articles that cover a number of related topics.
Source Databases. Source databases provide numerical data, a complete text, or a combination
of both. These include the many economic and financial databases and textual source databases
that contain the complete texts of newspaper or journal articles.
As opposed to the indices and summaries in the reference database, source databases provide complete textual or numerical information. They can be classified into (1) full-text information sources, (2) economic and financial statistical databases, and (3) online data and descriptive information on companies.
Lexis-Nexis has introduced three new services. Tracker scans thousands of publications daily and
delivers relevant news for only those topics designated by the customer. PubWatch allows users to
scan a particular publication table of contents and select only the stories they want to read. AM
News Brief provides news summary every day. Market research reports from more than a dozen
brand names such as Data Monitor, Find/SVP, and Nielsen are also available on Nexis.
In addition to the various major databases that provide financial information about companies
and stocks, such as Standard and Poor’s Compustat Services and the Value Line Database, a number of
online sources provide nonfinancial information about companies. Examples include the following:
110
Dun and Bradstreet Identifier: Over 57 million public and private companies, government agencies, and contractors, schools and universities with five or more employees, listing
addresses, products, sales executives, corporate organizations, subsidiaries, industry information, and sales prospects.
Disclosure: 12,500 U.S. public companies with at least $5 million in assets. Information
includes description of the business, balance sheet, cash flow, income statement, financial ratios, president’s letter, and management discussion.
American Business Directory: Over 10 million companies, mainly private. Also lists government offices and professionals, such as physicians and attorneys. Includes estimates of
sales and market share.
Standard and Poor’s Corporate Description Plus News: 12,000 public companies, with strategic
and financial information plus news. Includes business description, incorporation history, earnings and finances, capitalization summary, and stocks and bond data.
Invested Group MarkIntel: Comprises two databases, MarkIntel, which features reports authored
by top-rated business publishing firms, and MarkIntel Master, which features exclusive
online primary research from leading consulting and market research firms.
Data-Star: Has several individual files with full-text market research reports. Also provides a
Focus Market Research which includes Data Monitor, Euromonitor, ICC Keynote Report,
Investext, Frost and Sullivan, European Pharmaceutical Market Research, and Freedonia
Industry and Business Report.
NTIS: The National Technical Information Service is the official source for governmentsponsored U.S. and worldwide scientific, technical, engineering, and business-related
EXTERNAL SOURCES
OF
SECONDARY DATA
information. The NTIS collection includes business and management studies as well as
international marketing reports. The information is available in CD-ROMs, computer
tapes and diskettes, and online.
Classification of Databases Based on Storage and Retrieval Methods. Another useful way of
classifying databases is based on their method of storage and retrieval. They can be classified as
online databases and CD-ROM databases.
Online Databases. Online databases can be accessed in real time directly from the producers of
the databases or through a vendor. In order to access online databases, all one needs is a personal computer and Internet access. Online databases drastically reduce the time required for a
search and bring data right to the desk. Use of the Internet to obtain useful marketing research
information is discussed in Chapter 7.
CD-ROM Databases.5 Compact disk read-only memory (CD-ROM) technology has revolutionized the technology of storing and retrieving information. Storage and retrieval is no longer
restricted to large mainframe machines accessed by modems. Now, large amounts of information can be stored on compact disks and can be read by personal computers. A single CD-ROM
can hold approximately 600 megabytes of information (as much as 1,800 floppy disks) or 250,000plus pages of text. The main advantage of CD-ROM over online access is that there are no online
connect-time charges or long-distance telephone charges. The most powerful CD-ROM applications usually are sold by annual subscription or one-time fee for unlimited data access. Typically,
the user receives a disk with updated information each week, month, or quarter. Almost all the
reference and source databases that are available online are also available on CD-ROM.
Accessing Computer Retrievable Databases
Online databases are accessible both from their producers and increasingly from online information services. Most online services charge a fee for access to each database, a charge for
the amount of information retrieved, and possibly supplemental charges, depending on the
nature of the information or the contract arrangements. Some leading vendors of online databases are BRS (Bibliographic Retrieval Service), Knight-Ridder Information Service (formerly
DIALOG), Dow Jones News/Retrieval Service, Inc., and Mead Data Central. Commercial online
services such as America OnLine, Compuserve, and Prodigy also allow access to a number of
online databases. If you are a frequent online user, the vendors offer a number of options. One
way is to pay a monthly fee that entitles the user to a certain minimum usage and after that the
user is charged per minute. The actual cost of each service varies from vendor to vendor. In order
to access the information on a CD-ROM, it is necessary to have a CD-ROM drive connected to
a personal computer (or access to one available in a public library).
Advantages of Computer-Retrievable Methods
The main advantage of computer-retrievable methods is the scope of the information available
on databases. They now cover several thousand U.S. and worldwide information sources. A second advantage is the speed of information access and retrieval. Often, much of the information
is available from a computer before it is available in published form, because of the time
required for printing and mailing printed material. Third, commercially available search procedures provide considerable flexibility and efficiency in cross-referenced searching. For example,
by using the EIS Industrial Plants database, it is possible to locate plants that simultaneously
meet several criteria, such as geographic location, industry code, and market share. The future
of computer-retrievable databases is exciting. Marketing Research in Action 5-1 gives an idea
about the computer-retrieval databases of the future.
111
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
MARKETING RESEARCH IN ACTION 5-1
Virtual Reality
Virtual reality (VR) is being used for applications ranging from three-dimensional animation for games
to geophysical analysis. Virtual reality is a human–computer interface that does not include keyboard
and mouse; rather, the interface tracks the movement of the user’s hands and head. The movements of
the human body replace the clicking and typing. This not only helps to input data faster but also makes
the visualization of results very convenient.
VR tools include:
■ Visualization of all data types simultaneously in a common virtual world
■ Walk-through and fly-through capabilities
■ Multiattribute visualization
■ Interactive real-time region growing
■ Import and export of data
The benefits of these tools in data analysis are many. They allow a new user to understand the data
faster and thus be productive in a limited time frame. They also allow a user to easily understand complex relationships between data types and multiattribute visualization. Because interactive real-time
region growing is possible, there is improved understanding of various attributes that can affect the data.
The biggest advantage is improved interdisciplinary collaboration. By integrating different data types in
a common virtual work and by using large immersive screens, a group of people can share the same
experience and take an active part in the work in front of the screen.
SOURCE: “Inside Reality,” http://oilit.com/2journal/2article/0007/0002.pdf, January 2003.
Limitations of Computer-Retrievable Methods
The main limitations of the reference databases are their reliance on the accuracy of the abstract
author, the dependence on the journal and article selection policy of the database producer, and
the idiosyncrasies of the search procedures of the different databases as well as the different database network vendors.6
Because the computer search is based on finding certain key words within the abstract, there
is the possibility that some important information will be missed if an abstract is missing a key
word. On the other hand, a lot of irrelevant data may be generated if certain key words used to
limit a search are not cited in an abstract. For example, a manufacturer of minicomputers who is
interested only in developments pertaining to minicomputers may not want to retrieve the entire
database on computers. However, the abstract may contain the word “computer” regardless of size,
and accessing information on minicomputers would also yield general computer information.
Another limitation arises from the enormous amount of information now available online. It
is often quite difficult to know which of the myriad sources has the correct information most
readily accessible. Finally, the researcher using online database retrieval services must weigh the
benefits of the research procedure, including timeliness, speed, and scope of information
retrieval, against the costs of searching and accessing computer-retrievable databases.
CENSUS DATA
The demographic, economic, and social statistics contained in great detail in census data are key
aspects of many marketing studies, for example:
112
Company Y must decide where to locate a new shopping mall and which kinds of stores to install. These
decisions will require (1) census of population information about the populations with access to the proposed locations, (2) census of retail trade information about likely competitors and local wage levels, and
CENSUS DATA
(3) census of construction industries information about land development, contractors, and construction
costs, available by state and metropolitan area.
Understanding the Census
All countries conduct a mandatory enumeration of important facts about their population and
the economic and social environment. The major national and international census data collection agencies and some of their major publications are the U.S. Bureau of the Census (www.
census.gov/), Statistics Canada, Statistical Office of the European Communities (Social Statistics,
Industrial Statistics), Great Britain Central Statistical Office (Annual Abstract of Statistics), Japan
Bureau of Statistics (Japan Statistical Yearbook), and the United Nations (Statistical Yearbook).
The U.S. Bureau of the Census is illustrative of the scope of these undertakings. There are
actually eight regular economic censuses, taken in the years ending with the numbers 2 and 7, and
censuses of population and housing, which are taken every 10 years in the year ending with 0. The
eight economic censuses compile detailed statistics on the structure and functioning of the major
economic sectors: agriculture, construction industries, manufacturers, mineral industries, retail
trade, service industries, transportation, and wholesale trade.
Two major innovations were introduced in the 2000 Census. One is the availability of census data on CD-ROM. For a fee, the Census Bureau provides detailed summaries of the information it obtains on CD-ROM. The bureau also sells computer software that may be used for
accessing and tabulating data on the CD-ROM. The second major innovation in the 1990
Census was the introduction of the Topologically Integrated Geographic Coding and Referencing (TIGER) system. The TIGER system gives the user the ability to generate a digitized
street map of the entire 3.6-million-square-mile map of the United States. Specifically, with
use of the TIGER system, one can literally chart every block in every county in the United
States, both topographically and demographically. Five versions of the TIGER maps are available from the federal government. These include the prototype and precensus versions, both
issued in 1989; the Initial Voting Codes version, released in October 1990, which blankets the
United States by election districts; and the initial and final postcensus versions, the latter
released in the early part of 1991. TIGER covers 3,286 counties in the United States, including addresses from the most populated urban areas to the most rural areas. One of TIGER’s
most popular uses is to plot store locations.
To use census data effectively, one must be able to locate quickly the specific information relevant to the research topic. The Index to Selected 2000 Census Reports and the Index to 2000
Census Summary Tapes list all the titles of tables available from the 2000 Census in either printed
or tape form. Each table is described in terms of the variable and the level of aggregation used.
For example, one table may be described as “education by sex,” with an indication of the level
of aggregation available.
Census data can be obtained at many levels of aggregation (see Figure 5-3). The smallest
identifiable unit is the city block bounded by four streets and some other physical boundaries.
City blocks then are combined arbitrarily to form block groups. The block groups then are collected together to make up census tracts, which are generally used to approximate neighborhoods. Census tracts have populations of above 4,000 and are defined by local communities. In
urban areas, census tracts are combined to form metropolitan statistical areas (MSAs), which
are counties containing a central city with populations of at least 50,000.
The general concept of a metropolitan area is one of a large population nucleus, together
with adjacent communities that are determined to have a high degree of economic and social
integration with that central nucleus. A metropolitan statistical area containing a single core
with a population of at least 2.5 million may be subdivided to form smaller groupings of counties referred to as metropolitan divisions. While a metropolitan division is a subdivision of a
larger metropolitan statistical area, it often functions as a distinct social, economic, and cultural
area within the larger region.7 In June 1983 the federal government replaced the old standard
metropolitan statistical area (SMSA) designation with new definitions. To maintain comparability, data for an earlier period are revised where possible, to reflect the MSA boundaries of
113
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
County C
County A
City
Central City
Village
Block
group
County B
County D
Central City
Census tract
Elm
St.
Map
17th
16th
Ave.
Ave.
Urbanized
area
Block
le St
.
FIGURE 5-3
Geographic subdivisions of a MSA.
the more recent period. In addition to the new standard MSAs, the largest defined areas are
combined statistical areas (CSAs), which are metropolitan complexes containing separate component areas. [There are 367 MSAs and 68,000 census tracts and 211,267 block groups in the United
States (including Puerto Rico).]
Finally, the whole country is divided into four large regions (Northeast, Midwest, South,
and West). In addition, census data are available by civil divisions, such as states, counties,
cities, and wards. The data from the latest census can be obtained from the Census Bureaus
online database called Cendata. Cendata has an efficient measuring system that leads you to
the data you need.
NORTH AMERICAN INDUSTRY CLASSIFICATION
SYSTEM
The North American Industry Classification System (NAICS)8 is the new standard code system to describe business establishments and industries, replacing the Standard Industrial Classification (SIC) codes. It is the first industry classification system developed in accordance with
a single principle or aggregation: the principle that producing units that use similar production
processes should be grouped together in the classification. (See Table 5-2.)
TABLE 5-2
North American Industrial Classification System
NAICS level
NAICS code
Description
51
Information
Sector
Subsector
Industry group
Industry
U.S. industry
114
515
5151
51511
515112
SOURCE: http://www.census.gov/eos/www/naics/
Broadcasting (except Internet)
Radio and TV
Radio Broadcasting
Radio Stations
APPRAISING SECONDARY SOURCES
The NAICS went into effect for reference year 1997 in Canada and the United States and for
1998 in Mexico. It was developed to provide a consistent framework for the collection, analysis,
and dissemination of industrial statistics. Designed by the U.S., Mexican, and Canadian governments, NAICS has implications for economists, regulators, marketers, publishers, and anyone
else who uses industry-based data. The most obvious use of the codes is in the 1997 Economic
Census. Of the 1,170 NAICS codes, 358 are new industries, 390 are revised from SIC, and 422
can be compared to the older SIC codes.
The system was developed by the United States, Canada, and Mexico to provide comparable statistics across the three countries. For the first time, government and business analysts will
be able to directly compare industrial production statistics collected and published in the three
North American Free Trade Agreement countries. NAICS also provides for increased comparability with the International Standard Industrial Classification System (ISIC, Revision 3) developed and maintained by the United Nations.
NAICS industries are identified by a six-digit code, in contrast to the four-digit SIC code.
The longer code accommodates the larger number of sectors and allows more flexibility in designating subsectors. The international NAICS agreement fixes only the first five digits of the code.
The sixth digit, where used, identifies subdivisions of NAICS industries that accommodate user
needs in individual countries. There are 20 broad sectors in the NAICS system, up from the 10 divisions of the NAICS system.
APPRAISING SECONDARY SOURCES
Users of secondary sources rapidly develop a healthy skepticism. Unfortunately there are many
reasons why a forecast, historical statistic, or estimate may be found to be irrelevant or too inaccurate to be useful. Before such a judgment can be made, the researcher should have answers to
the following questions:
1. Who? This question applies especially to the reputation of the collecting agency for honest
and thorough work and the character of the sponsoring organization, which may influence
the interpretation and reporting of the data. A related question is whether either organization has adequate resources to do a proper job. The problems do not end here, for the
original data source (which provided the count, estimate, or other basis for the reported
result) may have its own motives for biasing what it reports. A company that is pressed
by a trade association, chamber of commerce, or government agency may be unwilling to
report the true state of affairs or to take the time to collect the data, which may result in
a biased guess.
2. Why? Data that are collected to further the interests of a particular group are especially suspect. Media buyers, for example, soon learn to be wary of studies of media. It is easy to choose
unconsciously those methods, questions, analysis procedures, and so forth, that favor the
interests of the study sponsor, and it is unlikely that unfavorable results will be exposed to
the public.
3. How? It is impossible to appraise the quality of secondary data without knowledge of the
methodology used to collect them. Therefore, one should immediately be suspicious of any
source that does not describe the procedures used—including a copy of the questionnaire (if
any), the nature and size of the sample, the response rate, the results of field validation efforts,
and any other procedural decisions that could influence the results. The crucial question is
whether any of these decisions could bias the results systematically.
The need for caution is illustrated by a study to determine the best locations for new bank
branches. The researchers initially used the projections of population in different parts of the
city, which were provided by the city planning commission. These ostensibly valuable data
had to be discarded when it was found that the commission had arrived at their projections
by subdividing on maps the areas to be developed and multiplying each area by the density
of families in the already-established areas of the city. When this methodology was discovered,
115
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
a proper projection was made by canvassing every real estate developer in the area regarding his or her future plans. The difference between the two projections was great, both in
extent and timing of population increases.
4. What? Even if the available data are of acceptable quality, they may prove difficult to use or
inadequate to the need. One irritating and prevalent problem is the classifications that are
used. Wide variations in geographic, age, and income groupings across studies are common;
for example, there is no accepted definition for the minimum number of stores in a supermarket chain.
5. When? There is nothing less interesting than last week’s newspaper. Sooner or later, the pace of
change in the world in general, and in markets in particular, renders all secondary data equally
obsolete and uninteresting except to the historian. The rate of obsolescence varies with the types
of data, but in all cases the researcher should know when the data were collected. There may
be a substantial lag between the time of collection and the publication of the results.
6. Consistency? With all the possible pitfalls in secondary data, and the difficulty in identifying
them fully, the best defense is to find another source that can be used as a basis for comparison. Ideally, the two sources should use different methodologies to arrive at the same kind
of data. In the likely event that there is some disagreement between the two sets of data, the
process of reconciliation should first identify the respective biases in order to narrow the differences and determine which set is the most credible.
APPLICATIONS OF SECONDARY DATA
Secondary data are widely used for a number of marketing research problems. We have already
discussed the various sources of secondary data and how to appraise them. In this section we
will look at the various applications of secondary data. Table 5-3 gives a comprehensive framework of the types of sources to be used for different applications.
Demand Estimation
Most marketing resources, especially sales effort, service coverage, and communication activity,
are allocated by region, segment, or territory. The key to efficient allocation is knowledge of the
potential of each segment relative to other segments. Hence, demand estimation is a key determinant of the allocation of resources. Demand can be estimated from secondary data by the methods described in the paragraphs that follow.
TABLE 5-3
Applications of Secondary Data
116
Demand Estimation
Monitoring the Environment
1. Census data
1. Press releases
2. North American Industry Classification System (NAICS)
2. Legislation and laws
3. Trade associations data
4. Experts and authorities
4. Business and practitioner literature, such
as magazines
Segmentation and Targeting
Developing a Business Intelligence System
1. PRIZM
1. Competitors’ annual reports
2. CLUSTER PLUS
2. Press releases
3. DMI
3. Blogs
4. NAICS
4. Competitors’ web pages
5. TIGER
3. Industry news
APPLICATIONS
OF
SECONDARY DATA
Direct Data Methods. Direct data methods are based on a desegregation of total industry data.
The sales information may come from government sources, industry surveys, or trade associations. For example, the National Electrical Manufacturers Association reports shipments of
refrigerators to retailers. Such data are useful for establishing relative market potentials only
if the sales can be broken down by the organization’s sales or operating territories. Fortunately,
industry refrigerator shipment data are available by trading area. This permits a direct comparison of the share of company sales and industry sales in each territory.
Corollary Data Methods. One solution to the absence of industry sales data for each territory
is to use another variable that is (1) available for each sales territory or region and (2) correlated
highly with the sales of the product. For example, the territory demand for child-care services
or baby food is correlated highly with the number of births in the area during the previous three
years. Thus, the share of births in all geographic areas within the territory of interest would be
a good proxy for the relative market potential within that territory.
Companies use other methods that are peculiar to their product and sales environment to
forecast sales. An example of such a method is given in Marketing Research in Action 5-2. In the
example, Hansen Company, a manufacturer of quick connective couplings for air and fluid
power transmission systems, uses a “sales-per-employee ratio” to forecast sales.
Monitoring the Environment
One of the most important uses of secondary data is to monitor the environment in which the
company is functioning. Monitoring the environment is very crucial these days, because it is
highly volatile and because attitudes, fashions, and fads change so often. To keep abreast of all
the latest developments, a company has to be in constant touch with newspapers, general magazines, and periodicals. It has to know all the latest legislation and laws that may affect it. To
know about the most recent trends in the industry, a firm has to look in the latest journals in the
field. Thus, constant monitoring of the environment through surveillance of all the relevant indicators is very important to compete effectively in this dynamic environment.
Segmentation and Targeting
Market segmentation is common among businesses seeking to improve their marketing efforts.
Effective segmentation demands that firms group their customers into relatively homogeneous
groups. The North American Industry Classification System (NAICS) and Dun’s Market Identifiers (DMI) are used by companies selling industrial goods to segment their markets.
One of the latest developments with regard to segmentation for consumer products is
geocoding, or a cluster demographic system, which identifies groups of consumers who share
demographic and lifestyle characteristics. Several services now can link U.S. Census data on a
Zip Code basis to lifestyles, to help marketing researchers identify the best areas in which to
concentrate their efforts. Among the services are PRIZM (by Claritas) and CLUSTER PLUS (by
Donnelly Marketing Information Services).
About 1,000 consumer characteristics can be used to build clusters of homogeneous groups.
Demographic variables include age, marital status, size of household, and income. Marketing
Research in Action 5-3 illustrates how important demographic variables are in segmentation and
targeting. Behavioral characteristics include amount of TV watched, amount of white bread consumed, types of magazines read, and so on. Geographic areas range from cities and counties to Zip
Codes, census tracts, and block groups.
The Potential Rating Index Zip Markets (PRIZM) system is based on evidence that people
with similar cultural backgrounds and circumstances will gravitate naturally toward one another.
Each of the 35,600 zip markets was first described according to 34 key demographic factors. These
zip markets were originally clustered into 40 distinct groups which were each very homogeneous
within themselves and very different from other groups. Subsequently 22 new consumer groups
were added, bringing the total number of clusters to 62. The larger number of clusters reflect the
117
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
MARKETING RESEARCH IN ACTION 5-2
Relative Sales Potentials
What can be done if a reliable estimate of total industry sales is unavailable, the customers cannot provide
a good estimate of their purchases of the product, and the product is used in many industries, so there are
no obvious corollary variables? This was the situation confronting the Hansen Company, a manufacturer
of quick connective couplings for air and fluid power transmission systems, who distributed these products through a national network of 31 industrial distributors. To be able to evaluate and control their activity, the company badly needed data on the relative performance of their distributors. Its approach was
based on the only reliable data that were available to it—sales of company products, by distributor. To utilize these data, the company made the assumption that it should be possible for Hansen distributors to
attain the same sales-per-employee ratio in noncustomer establishments as in customer establishments. To
establish the sales-per-employee ratio that would serve as a performance standard, the following steps
were taken:
1. A random sample of 178 accounts was drawn from a census of all customer accounts buying $2,000
or more from the seven best distributors (where “best” was defined in terms of perceived effectiveness of management and utilization of an up-to-date data processing system).
2. Each account was assigned to a two-digit SIC group on the basis of its principal output or activity.
3. Data on the number of employees in each account were obtained primarily from industrial directories.
4. Sales-per-employee ratios were computed for each SIC group within the set of seven distributors.
5. Sales-per-employee ratios for each SIC group were multiplied by the total employment in all establishments in each of the 31 distributor territories. The employment data came from the current edition of the County Business Patterns publication of the U.S. Census. The output of these five steps
was a table for each distributor, patterned after the following table, which gives the results for distributor A.
1997 Sales Potential—Distributor A
Two-Digit SIC Group
1997 Hansen Sales
per Employee
33
$1.56
4,113
34
2.53
14,792
37,424
35
3.28
15,907
52,175
36
1.93
32,677
63,067
37
1.71
2,024
3,461
38
3.48
409
1,423
69,922
$163,966
Total
Total Employees
Sales Potential
$
6,416
The resulting sales potential was compared with actual sales, which for distributor A amounted to
$86,218 in 1997. That is, actual sales performance was 52.6 percent of sales potential. The sales performance for all distributors ranged from 125.0 percent to 15.4 percent, with an average of 50.3 percent.
It is not surprising that the distributor with sales of only 15.4 percent of potential was subjected to a
very careful review, which revealed that the salespeople did not really know how to sell the product to
major accounts in the area.
118
increasing ethnic and economic complexity of the nation’s population.9 Claritas has also introduced workplace PRIZM, which accurately profiles a market’s working population and demonstrates the difference between the area’s daytime and nighttime demographics.
Since every market is composed of Zip Code areas, it is possible to estimate the sales potential of a market by zip market clusters. As an example, a power tool manufacturer was able to
create a PRIZM profile of product warranty cards mailed by recent buyers. This told the manufacturer which Zip Code areas should be chosen as target markets and helped to allocate media
spending and sales force effort.
SOURCES
OF
SECONDARY DATA
FOR INTERNATIONAL
MARKETING RESEARCH
MARKETING RESEARCH IN ACTION 5-3
Targeted Marketing for African-Americans
African-Americans comprise of about 13 percent of the population in the U.S. (around 39.2 million) and
by 2050, this number is predicted to rise by 15 percent. With their spending on products and services
being on the increasing trend, it provides enough incentive for companies to target this segment and
come up with innovative ways to reach them. Marketers adopt various strategies to win over a particular segment of the market, some of which are given below:
■ Segmentation – It is critical for a company to divide its target audience into segments based on ethnicity, lifestyle and other demographics like education, household income etc. Each of these segments must be analyzed in-depth and the company must look for inventive ways to reach out to
them. A good example of this is the case of Nissan. Research done by the company revealed that
their African-American consumers preferred to be ‘included’ rather than ‘targeted’. This led Nissan
to embark on a series of ‘integrated marketing’ for their Altima brand which did not directly target
the African-American population but, subtly included them in the target audience. Through this initiative, Nissan saw a 14 percent increase in their sales of Altima to African-American audience and
a 29 percent overall increase in African-American customer base.
■ Community-based events – Another tested method of gaining attention among segments in the market
is hosting community based events such as exhibitions etc. New York-based Merrill Lynch & Company
Inc has conducted many of these events and benefited greatly from them. In October 2005, the firm
sponsored a cultural exhibition: The Art of Architecture in Africa, and invited more than 300 prospective clients who were attended to by over 25 financial advisors from Merrill Lynch.
There has been increasing focus on the African-American consumer segment. Some of the successful
industries in this initiative are automotive manufacturers, financial services, wireless providers, consumer packaged goods and insurance companies. The key to succeed in this multicultural marketing
campaign lies in a companywide commitment to ensure success, cultural sensitivity and the ability to
capture their attention and loyalty.
SOURCE: Deborah L. Vence, “Mix it up,” Marketing News, October 15, 2006.
Developing a Business Intelligence System
A business intelligence system is basically a system that contains data on the environment and
the competitors. It forms an integral part of the marketing decision support system. Both primary and secondary data form a part of the business intelligence system. As has already been
said, data on the environment can be obtained from a variety of sources. Data on competitors
can be obtained from their annual reports, press releases, patents, and so on.
SOURCES OF SECONDARY DATA FOR
INTERNATIONAL MARKETING RESEARCH 10
Secondary data are a key source of information for conducting international marketing research.
This is in part due to their ready availability, the high cost of collecting primary data versus the
relatively low cost of secondary data, and the usefulness of secondary data in assessing whether
specific problems need to be investigated, and if so, how. Further, secondary data sources are
particularly valuable in assessing opportunities in countries with which management has little
familiarity, and in product markets at an early stage of market development.
A wide variety of secondary data sources are available for international marketing research.
These range from sources that provide general economic, social, and demographic data for
almost all countries in the world, to sources that focus on specific industries worldwide.
A host of sources of macroeconomic data are to be found, ranging widely in the number of
countries or regions covered. Many of these are based on or derived from United Nations and
119
CHAPTER 5 SECONDARY SOURCES
OF
MARKETING DATA
World Bank data. The Business International, Euromonitor, and Worldcasts divisions of Predicasts also publish annual information on macroeconomic variables.
The preceding macroeconomic data sources, with the exception of Euromonitor, relate to the general business environment. They therefore do not provide much indication as to market potential for
specific industries. A number of sources of industry-specific data are available. They are United
Nations Yearbooks, publications of the U.S. Department of Commerce, The Economist, and the Worldcasts.
Numerous other sources specific to individual countries or product markets are also to be
found. The U.S. Department of Commerce, for example, publishes International Marketing Handbook, which provides profiles and special information about doing business in various countries.
Information regarding regulations, customs, distribution channels, transportation, advertising
and marketing research, credit, taxation, guidance for business travelers abroad, and so forth, is
compiled in their “Overseas Business Reports.” Governments or other bodies frequently publish
national yearbooks or statistical data books. Various private sources also publish regional and
country handbooks. The World of Information, for example, publishes the African Guide, the
Middle East Review, and so on.
PROBLEMS ASSOCIATED WITH SECONDARY
DATA IN INTERNATIONAL RESEARCH
Two major problems are associated with secondary data in international marketing research: the
accuracy of the data and the comparability of data obtained from different countries.
Data Accuracy
Different sources often report different values for the same macroeconomic factor, such as gross
national product, per-capita income, or the number of television sets in use. This casts some
doubt on the accuracy of the data. This may be due to different definitions followed for each of
those statistics in different countries. The accuracy of data also varies from one country to
another. Data from highly industrialized nations are likely to have a higher level of accuracy
than data from developing countries, because of the difference in the sophistication of the procedures adopted. The level of literacy in a country also plays a role in the accuracy of the macroeconomic data collected in that country.
Comparability of Data
Business statistics and income data vary from country to country because different countries have
different tax structures and different levels of taxation. Hence, it may not be useful to compare
these statistics across countries. Population censuses may not only be inaccurate, they also may
vary in frequency and the year in which they are collected. Although in the United States they
are collected once every 10 years, in Bolivia there was a 25-year gap between two censuses. So
most population figures are based on estimates of growth that may not be accurate and comparable. Measurement units are not necessarily equivalent from country to country. For example,
in Germany the expense incurred on buying a television would be classified as entertainment
expense, whereas in the United States it would be classified as furniture expense.
APPLICATIONS OF SECONDARY DATA
IN INTERNATIONAL RESEARCH
120
Secondary data are particularly useful in evaluating country or market environments,
whether in making initial market-entry decisions or in attempting to assess future trends and
developments. They thus form an integral form of the international marketing research
SUMMARY
process. More specifically, three major uses of secondary data are in:
1. Selecting countries or markets that merit in-depth investigation
2. Making an initial estimate of demand potential in a given country or a set of countries
3. Monitoring environmental changes
Secondary data can be used systematically to screen market potential, risks, and likely costs
of operating in different countries throughout the world. Two types of generalized procedures
are used. The first procedure classifies countries on two dimensions: the degree of demographic
and economic mobility, and the country’s domestic stability and cohesion. The second procedure
calculates multiple factor indices for different countries. For example, Business International publishes information each year on three indices showing (1) market growth, (2) market intensity, and
(3) market size, for countries in Western and Eastern Europe, the Middle East, Latin America, Asia,
Africa, and Australia. Customized models, which are geared to specific company objectives and
industry characteristics, can also be developed using secondary data.
Once the appropriate countries and markets to be investigated in depth have been determined, the next step is to make an explicit evaluation of demand in those countries or markets.11
This is important when considering initial market entry, because of the high costs and uncertainty associated with entering new markets. Management has to make an initial estimate of
demand potential, and also project future market trends.
Four types of data analyses are unique to demand estimation in an international context. The
first and the most simplistic is lead–lag analysis. This uses time-series (yearly) data from a country to project sales in other countries. A second procedure is the use of surrogate indicators. This
is similar to the use of general macroindicators, but develops the macroindicators relative to a
specific industry or product market. An example of a surrogate indicator is the number of childbirths in the country as an indicator of the demand potential for diapers. A third technique, which
relies on the use of cross-sectional data (data from different countries), is analogous to the use
of barometric procedures in domestic sales forecasting. One assumes that if there is a direct relationship between the consumption of a product, service, or commodity and an indicator in one
country, the same relationship will hold in other countries to estimate the demand. The fourth
and most complex forecasting model is the econometric forecasting model. This model uses crosssectional and time-series data on factors underlying sales for a given product market for a number of countries to estimate certain parameters. Later, these models can be used to project the
market demand.
A third use of secondary data in an international context is to monitor environmental changes.
Monitoring environmental changes requires surveillance of a number of key indicators. These
should be carefully selected and tailored to the specific product or range of products with which
management is concerned. Two types of indicators are required. The first monitors the general
health and growth of a country and its economy and society; the second, those of a specific industry or product market. A variety of procedures can be used to analyze the impact of environmental factors on world trends or industrial countries, and on product markets, as well as the
implications for market growth and appropriate marketing strategies. These range from simple
trend projections or tracking studies and the use of leading indicators to the more complex scenario evaluation studies.
Summary
The theme of this chapter is the wealth of data available to marketing researchers. Many management problems can be resolved by recourse to the firm’s internal records or to secondary sources such as government
statistics, trade association reports, periodicals, books, and private studies. With the growing power of computers, these data are increasingly easy to access in databases. The low cost and convenience of these database sources leave no excuse for not starting a marketing research study with a thorough scan of what is
121
already available. Invariably, the researcher will be surprised at the extent of what is already available with
very little effort. Even if it is not entirely suitable, the secondary data sources can provide useful pointers
on how to design a good research study.
Questions and Problems
1.
2.
3.
4.
5.
You are opening a new retail store that will sell personal computers and software. What secondary data
are available in your area to help you decide where
to locate the store? Would the same data be relevant
to someone opening a convenience copying center?
A large chain of building supply yards was aiming
to grow at a rate of three new yards per year. From
past experience, this meant carefully reviewing as
many as 20 or 30 possible locations. You have been
assigned the task of making this process more systematic. The first step is to specify the types of
secondary information that should be available for
the market area of each location. The second step
is to identify the possible sources of this information and appraise their usefulness. From studies of
the patrons of the present yards, you know that 60
percent of the dollar volume is accounted for by
building contractors and tradesmen. The rest of the
volume is sold to farmers, householders, and hobbyists. However, the sales to do-it-yourselfers have
been noticeably increasing. About 75 percent of the
sales were lumber and building materials, although
appliances, garden supplies, and home entertainment systems are expected to grow in importance.
For each of these products, which industry associations would you contact for secondary data?
(a) Foreign convenience dinners, (b) numerically
controlled machine tools, (c) irrigation pipe,
(d) imported wine, (e) compact disk players, and
(f) children’s shoes.
Obtain data on beer consumption in your state or
province for the latest available year. Calculate the
per-capita consumption for this area and compare
it to that for the country as a whole. What accounts
for the difference?
Educational Edge, a small company with limited
resources, is interested in segmenting potential
markets for its erasable transparencies.
6.
7.
a. Which type of data would be best suited to
obtain the required information?
b. What are the possible sources of information to
aid in the segmentation decision?
c. What are the benefits and limitations of using
secondary data for this purpose?
Howard Enterprises, a small family-owned manufacturer of unique lamps, has begun to receive
unsolicited inquiries about its product from foreign countries. The company has been operating
exclusively in the domestic environment, but these
inquiries have become numerous enough to suggest that a market for these specialty lamps may
exist abroad. J. P. Howard, the company head,
decided to contact Peter Franks, an old college
friend of his, who is now the head of marketing
research for a multinational company, to ask for his
advice on how to proceed in evaluating foreign
country markets. Mr. Franks recommends that
Howard Enterprises should select countries that
merit in-depth investigation and proceed to make
an initial estimate of the demand potential in these
countries.
a. Considering the limited resources that are available to Mr. Howard’s company, explain how
secondary data can be used to help Mr. Howard
follow his friend’s recommendation.
b. What are the possible limitations of secondary
data of which Mr. Howard must be aware when
conducting the marketing research?
From secondary data sources, obtain sales for an
entire industry and the sales of the major firms
in that industry for any year. Compute the market shares of each major firm. Using another
source, obtain information on the market shares
of these same firms. Are there differences? If so,
why?
End Notes
1. Ellen R. Kidd, “Establishing Quality Focus in a
Multi-Cultural Organization,” presented at the
Third Congress on Competitive Strategies.
2. Paula A. Francese and Leo M. Renaghan, “Finding
the Customer,” American Demographics, January
1991, pp. 48–51.
3. More extensive discussion of data sources and how
to locate them can be found in Lorna Daniels,
122
Business Information Sources, Berkeley, CA: University of California Press, 1985; Lorna Daniels, Notes
on Sources of External Marketing Data, in B.
Shapiro, R. Dolan, and J. Quelch (eds.), Marketing
Management Vol. ii, Homewood, IL: Richard D.
Irwin, 1985, Appendix; Barbara E. Brown, Canadian
Business and Economics: A Guide to Sources of Information, Ottawa: Canadian Library Association,
ge / aak
w
om/ c o l
er
.c
le
w. wi l ey
1984; and Leonard M. Fuld, Competitor Intelligence,
New York: John Wiley, 1985.
4. A comprehensive listing of these associations can
be found in the Encyclopedia of Associations, Detroit:
National Gale Research Co., 1984; and in Leonard
M. Fuld, Competitor Intelligence, New York: John
Wiley, 1985.
5. For more information, see Jennifer Langlois, CDROM 1992: An Annotated Bibliography of Resources,
Westport, CT: Meckler, 1992.
6. An interesting compilation of reasons why a database search might not meet with success is provided
in Jeff Pemberton, “Faults and Failures—25 Ways
CASE 5-1
Barkley Foods
w
Joyce Stevenson, the manager of marketing research for
Barkley Foods, had just left an emergency meeting with
the firm’s president. An opportunity to buy an established line of gourmet (high-quality/high-priced)
frozen dinners had arisen. Because there were other
interested buyers, a decision had to be made within
three or four weeks. This decision depended on judgments about the future prospects of the gourmet frozen
dinner market and whether Barkley could achieve a
competitive advantage. The marketing research group
was asked to provide as much useful information as
possible within a 10-day period. Although uncomfortable with the time pressure involved, Joyce was pleased
that marketing had finally been asked to participate in
the analysis of acquisition prospects. She had pressed
for such participation and now she had to deliver.
Because of prior work on frozen fruit juices, Joyce
had some knowledge of the gourmet frozen market.
It was pioneered by Stouffer, who introduced the
Lean Cuisine line of entrees in 1981. Since then, other
firms have entered the industry with complete gourmet
dinners (including Swanson’s Le Menu and Armour’s
Dinner Classics). The distinction between entrees,
dinners, and the three main types of food offered—
conventional, ethnic (i.e., Benihana Restaurant Classics),
or low-calorie (i.e., Weight Watchers or Light &
Elegant)—define relevant submarkets. Joyce hypothesized that the gourmet frozen food buyer differs from the
buyer of conventional “TV dinners” in several respects.
The gourmet frozen food buyers are generally young,
upper-socioeconomic-group people who probably have
microwaves, are more health conscious, and are likely to
be working women and others who want sophisticated
cuisine but lack the time to prepare it.
Barkley Foods was a diversified food company
with sales of $2.3 billion. Over 80 percent of its sales
came from branded packaged food products sold
nationally through grocery stores. Its largest product
7.
8.
9.
10.
11.
That Online Searching Can Let You Down,” Online,
September 1983.
http://proximityone.com/metros.htm
http://www.census.gov/eos/www/naics/
Christina Del Valle, “They Know Where You Live—
And How You Buy,” Business Week, February 7, 1994.
For a more detailed discussion, see Susan P. Douglas and C. Samuel Craig, International Marketing
Research, Englewood Cliffs, NJ: Prentice-Hall, 1983.
V. Kumar, A. Stam, and E. A. Joachimsthaler, “An
Interactive Multicriteria Approach to Identifying
Potential Foreign Markets,” Journal of International
Marketing, 2(1), 1994, pp. 29–52.
areas were canned tomato products, frozen orange
juice, cake mixes, and yogurt. Barkley was known to
have strengths in operations (product preparation), distribution (obtaining distribution and managing the
shelves), and advertising. Their brands typically held a
solid second-place position in the supermarket. There
was no effort at umbrella brand identification, so each
product area was carried by its own brand.
Joyce Stevenson had previously been in strategic
planning, and reviewed the type of information and
analysis that would be required to support a strategic decision like this one. She wrote down the following four sets
of questions to guide the thinking of the research group:
1. Market analysis
■ What are the size, current growth rate, and projected growth rate of the industry and its relevant
subsets (such as ethnic dinners) for the next five
and ten years?
■ What are the important industry trends?
■ What are the emerging production technologies?
■ What are the distribution trends?
■ What are current and future success factors
(a competitive skill or asset needed to compete
successfully)?
2. Environmental analysis
■ What demographic, cultural, economic, or governmental trends or events could create strategic
threats or opportunities?
■ What major environmental scenarios (plausible
stories about the future) can be conceived?
3. Customer analysis
■ What are the major segments?
■ What are their motivations and unmet needs?
4. Competitor analysis
■ Who are the existing and potential competitors?
■ What are their current or forecasted levels of
sales, market shares, and profits?
■ What are their strengths and weaknesses?
■ What strategies are they following, and how are
they differentiating themselves in the market?
123
Questions for Discussion
1.
What secondary data sources would be useful? What
types of questions might be answered by each?
Identify one piece of information from the library
that would be helpful and relevant. How did you
locate it?
om/ c o l
ge / aak
w
.c
le
w. wi l ey
2.
CASE 5-2
Dell in Latin America?
er
w
Dell, which thrived while other PC makers stumbled in
1998, reported a 53 percent jump in profit and a 50 percent jump in revenue for its fiscal second quarter, which
ended in July 1998. Unlike most of its rivals, Dell deals
directly with customers and builds PCs only after
receiving an order. However, all is not well for Dell.
While Dell continues to blow away the competition and
Wall Street with enormous increases in personalcomputer sales, it is continually scrambling to bring in
and train enough people to keep up with its orders.
The company’s employment had grown 56 percent
in 1997, to 20,800, and within a span of three months
Dell added 225 people a week—about the same it
added every six weeks in 1996. To manage this
expansion Dell has aggressively recruited experienced
outsiders and tried to standardize training of new
employees. It has also created a culture where
3.
What other mechanisms would you use to gather
information?
managers are rewarded for seeing their divisions split
into smaller units and their responsibilities cut back.
Incidentally, Dell Computer Corp. is expected to
announce a major foray into Brazil, in a bid to boost its
share of Latin America’s fast-growing personalcomputer market. Foreign companies with local production plants dominate Latin America’s estimated $6.5
billion personal-computer market. Compaq Computer
Corp., International Business Machines Corp., Acer
Inc., and Hewlett-Packard Co. together accounted for
42 percent of desktop and notebook sales in 1997,
according to IDC Latin America, a market-research
firm. Dell ranked ninth with 1.2 percent of the market.
At the company’s annual meeting in July 1998, Vice
Chairman Morton Topfer said Dell aimed to open a
plant in Latin America in 1999 and noted that Latin
America is key to its international expansion. The company already operates regional production plants in
Malaysia, Ireland, the United States, and China.
Questions for Discussion
1.
2.
3.
What are the issues concerning the Latin American
market that Dell should address before it enters
into Brazil?
What kind of information is needed to address the
issues identified in question 1? What are the possible sources of the required information?
What are your recommendations to Dell regarding
the steps for expanding its operations in Latin
America?
CASE 5-3
Eddie Bauer: Strategize with
Secondary Marketing Data
Eddie Bauer, Inc., is a leading international retail brand,
offering casual lifestyle products for adults through its
retailing concepts: Eddie Bauer and Eddie Bauer Home.
Eddie Bauer products are for men and women who seek
versatile, classically styled, high-quality merchandise
124
Source: This case was prepared by V. Kumar and
Rajkumar Venkatesan for the purpose of
classroom discussion from “Dell to Build Plant to
Boost Latin Presence,” Wall Street Journal,
August 18, 1998, and “Dell Scrambles to Find Enough
Workers to Fuel Growth,” Wall Street Journal,
August 20, 1998.
designed to meet a wide range of their apparel and
home-furnishing needs. Since 1920, the company has
evolved from a single store in Seattle to a multichannel, international company with more than 590 stores
worldwide. In addition, Eddie Bauer has 110 million
catalogs in circulation and an online website
(www.eddiebauer.com). The company operates stores
in the United States and Canada, and has joint venture
partnerships in Germany and Japan.
Eddie Bauer has long considered the location of
stores as the key element for maximizing sales. However, the problem has been how to choose these prime
locations. Until recently, the company relied on basic
census information, such as income and number of
households, when choosing locations for new stores.
But recent demographic changes in the United States
and Canada have made this task more complex.
One of the main objectives of Eddie Bauer’s marketing team was to answer the following questions:
■ Who are our best prospects? What are their demographics: age, income, occupation, education, household size, and more?
■ What are they like? How would target customers like
to spend their leisure time, what kinds of products
do they purchase, what types of vehicles are they
likely to drive, and what is their interest in and use
of the latest technologies.
■ Where can I find them? What are the right places to
look—rural, urban, inner-city, small-town, or suburban communities? The next step is to figure out a
way to get down to the blocks they live on.
■ How can I reach them? How can customers be
attracted into Eddie Bauer stores and what are their
media habits: television, radio, newspapers, magazines, or the Internet?
So Eddie Bauer purchased Clarita’s online geocoding and GIS mapping software, along with databases
to provide precise knowledge for identifying ideal location. The first step was to identify potential areas of
high retail expenditures. Information (household
growth rate, number of housing units) about these locations along with consumer information (age, income
level, household size, and education level) were collected using Clarita’s demographics. Thus, for each
proposed store location, Eddie Bauer’s marketing team
gathered detailed consumer information on the market
and potential new locations relative to existing stores
and competitors.
Moreover, Eddie Bauer’s marketing team started to
use Clarita’s Potential Rating Index Zip Markets
(PRIZM) geocoding system to evaluate existing sites
and current customers. For example, they not only
profiled the shoppers at its five best-performing retail
outlets but also studied the branches that were not
doing very well. The profile of the shoppers would help
the marketing team identify the shared characteristics of
customers who frequent Eddie Bauer Stores. This
knowledge would be very helpful not only in deciding
on new site locations, interior designs, and merchandising mix but also in developing the store concept. This
detailed consumer information prepares Eddie Bauer to
compete more efficiently in the ever-changing retail
industry.
Questions for Discussion
1.
2.
3.
What demographic and geographic information
should Eddie Bauer collect to select store locations?
What kind of information regarding the competitive environment of the market in which Eddie
Bauer operates is needed when choosing store
locations?
Suggest methods by which Eddie Bauer can conduct marketing research to identify the shared
characteristics of those customers who frequent
Eddie Bauer stores versus the characteristics of
customers who frequent the stores of competitors
at various locations.
4.
Look at Clarita’s website (http://www.nielsen.com/
us/en.html) and try to find answers to the four
questions of Eddie Bauer’s marketing team: Who
are our best prospects? What are they like? Where can
I find them? How can I reach them?
Source: This case was prepared by the authors
with the inputs from Eddie Bauer, Inc.
(www.eddiebauer.com) and Claritas, Inc.
(www.claritas.com).
125
6
STANDARDIZED SOURCES
OF MARKETING DATA
Learning Objectives
■ Introduce the various sources of standardized sources of marketing data available.
■ Provide a description of each of the well-known sources of standardized marketing data.
■ Discuss the sources and applications of scanner data.
■ Develop a framework for the various applications of standardized data sources.
Marketing intelligence is a form of business intelligence whereby data and information are analyzed and transformed for use in strategic planning and problem solving. The more specific and
topical the need for information, the smaller is the likelihood that relevant secondary data will
be found. The researcher then has the choice of designing a special study or taking advantage
of standardized data collection and analysis procedures. The latter alternative generally exists
whenever several information users have common information needs and when the cost of satisfying an individual user’s need is prohibitive. These conditions are most often encountered
with consumer goods sold to large, diffuse markets and repurchased at frequent intervals. A further condition—especially important for data sources such as store audits and continuous consumer panels—is that the information needs are recurrent and can be anticipated. Thus, the data
supplier can enter into long-term relationships with clients and be sure of covering the heavy
fixed costs. The clients get continuity of data series, which is essential for monitoring and evaluation purposes.
This chapter describes and evaluates the major syndicated sources of marketing data, including store audits, warehouse withdrawal services, consumer purchase panels, and scanner-based
systems. Each source has a distinctive profile of strengths and weaknesses that reflects differences in orientation, types of measures, and their location in the distribution channel. To get a
full picture of the market situation of a product category or brand, it is usually necessary to use
several sources in combination. Unfortunately, when this is done the result is more often confusion rather than clarity of insight, because of information overload. This is such a prevalent
problem that the last section of this chapter is devoted to recently developed decision support
systems that help reduce the confusion.
The use of standardized data sources has been revolutionized by single-source data from
scanner systems. This means that all data on product purchases and causal factors, such as media
exposure, promotion influence, and consumer characteristics, come from the same households.
These data are being made possible through advances in information technology. Single-source
data has not fully displaced other standardized sources, but it is used in conjunction with them
to generate important new insights.
To understand the basic motivation for using standardized sources, consider the problems
of a manufacturer of cold remedies who has to rely on factory shipment data for sales information. Management is especially interested in the reaction to a new convenience package that was
introduced at the beginning of the cold season in December. By the end of January, the following information had been received from the accounting department:
126
CONSUMER PURCHASE PANELS
Week Ending
December 28
January 4
January 11
January 18
January 25
Factory Shipments
12,700 cases
19,800 cases
18,200 cases
14,100 cases
11,050 cases
All the usual problems of interpreting time-series data are compounded in this example by
the ambiguities in the data. For example, while the cases were shipped from the factory, there is
no information concerning where or when they were sold. The peak in factory shipments during the week ending January 4 represents a substantial amount of “pipeline filling,” in anticipation of sales to come in the future during the cold season, and an unknown amount of product
sold for pipeline filling still remains on the shelves. Also unknown is competitive performance
during this period: Did the new package gain sales at the expense of competition, or was there
a loss of market share? Shipment data provide no diagnostic information, so these questions
remain: How many retail stores used the special displays of the new product? Were competitors
making similar or more effective offers? Was there a carryover of last year’s stock in the old package? Without answers to these questions, the manager is in no position either to correct problems or continue the strong points of the campaign.
RETAIL STORE AUDITS
Every two months a team of auditors from a research firm visits a sample of stores to count the
inventory on hand and record deliveries to the store since the last visit. Sales during the twomonth period, for any desired classification of product category (including brands, sizes, package types, flavors, etc.), are arrived at by computing
Beginning inventory ⫹ deliveries ⫺ ending inventory ⫽ sales for the period
These retail store audit results then are projected—to arrive at nationwide and regional estimates
of total sales, inventories, and so forth—and reported to the client between six and eight weeks
after the end of the period. During each store visit, the auditor also may collect such observable
information as shelf prices, display space, the presence of special displays, and in-store promotion activities.
Nielsen Retail Index
A.C. Nielsen Co. and Audits and Surveys provide auditing services. Nielsen’s auditing services
cover four major reporting groups:
1. Grocery products
2. Drugs
3. Other merchandise
4. Alcoholic beverages
Audits and Surveys deals with products, regardless of the type of retail outlet carrying the product. The data provided by these two companies are incredibly rich. Table 6-1 summarizes the
information that is provided in the bimonthly report of Nielsen. A disadvantage of the audit data
is that it cannot be linked to household characteristics. For additional information, visit the A.C.
Nielsen website at http://ca.nielsen.com/content/nielsen/en_ca/insights/reports.html.
CONSUMER PURCHASE PANELS
From store audits and warehouse withdrawal services, we can learn how much product is moving through the distribution channel. As this information is one step removed from the actual
purchase transaction, we still don’t know who bought, how frequently they bought, or whether
127
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
TABLE 6-1
Contents of a Nielsen Store Audit Report
Each
a.
b.
c.
d.
of the following variables can be subdivided as follows:
Sales districts
Size of county (A, B, C, or D)
Type of store (for example, chain versus large-medium-small independents)
Thirty-two largest metropolitan markets
1. Sales (volume, trend, and share) on the basis of retail dollars and units, pounds or equivalent cases
for total market, and major brands by sizes, flavors, types, etc., as appropriate to the category
2. Distribution
a. Percentage of all stores, and all commodity sales, carrying each brand, and size
b. Out-of-stock conditions
c. Retail inventories
d. Stock cover (the length of time the stocks will last, assuming current rates of sales)
e. Source of delivery (wholesaler, rack jobbers, manufacturer, chain warehouse, or interstore
transfers)
3. Selling prices and volume sold at each price or deal
4. Retailer support in terms of shelf facings, special displays, in-store advertising, and newspaper
advertising
5. Media advertising for total market and major brands
6. Special analyses (illustrative)
a. Analyses of combinations of brands stocked to determine the extent to which individual brands
compete together
b. Cumulative distribution of new products
the seeming stability of market shares reflects stable purchasing patterns or a great deal of switching back and forth between brands and stores in response to short-term promotional efforts. To
answer these questions, we need detailed records of purchasing activity by the same people over
an extended period of time. Here are two methods for collecting this data:
1. In the home audit approach the panel member agrees to permit an auditor to check the
household stocks of certain product categories at regular intervals. A secondary condition is
that the panel member save all used cartons, wrappers, and so on, so the auditor can record
them.
2. In the mail diary method the panel member records the details of each purchase in certain categories and returns the completed diary by mail at regular intervals (biweekly or monthly). The
detail that can be collected is illustrated for 2 of the 88 clothing, food, and personal-care products recorded by members of the American Shoppers Panel (see Figure 6-1).
128
Both types of panels are used extensively in Europe, whereas in the United States and
Canada the mail diary method is dominant. When comparisons have been possible, the two
methods have produced equally accurate market share and trend data.1
National Family Opinion (NFO) and the National Purchase Diary Panel (NPD) both operate
mail diary panels.
Each month MRCA delivers a report for the preceding month, showing total consumer purchases of each brand in the category, plus sales overall, which can be converted into volume and
dollar market shares. Each quarter, an analysis of purchases by regions and by type of retail outlet is provided.
The IPSOS Home Testing Institute (HTI) Consumer Panel gives you a direct link to 350,000
households (almost a million individuals) who have agreed to participate in IPSOS studies conducted by mail, phone, or the Internet. Within the panel, a core sample of 250,000 households is
geographically and demographically balanced to provide an accurate picture of today’s marketplace.
NOTE: Extra space page 1. Free Gifts and Samples, inside front cover.
SEX
AGE
AGE
AGE
SEX
No. 6
AGE
User
No. 5
SEX
User
No. 4
SEX
User
No. 3
SEX
User
No. 2
AGE
SEX
Sex
AGE
THINS
OTHER
PLATINUM
TUNGSTEN
CHROMIUM
BLUE
SUPER BLUE
STAINLESS STEEL
OTHER
BONDED
TWIN (2) BLADED
CARTRIDGE
INJECTOR
CONTINUOUS
EDGE (BAND)
SINGLE EDGE
Age
AGE AND SEX OF USERS
User
Check (√)
one
LOW CALORIE
CAN
WRITE IN FLAVOR
Such as: Orange, Chocolate,
Cola, Lemon, Cherry,
etc.
COPY FROM LABEL
REGULAR
OTHER
CASE
Other–
Describe
PLASTIC BOTTLE
¢
Total
¢ off
WRITE IN AS MANY AS APPLY TO THE ITEM,
WRITE IN
AGE, SEX
OF
User
PURCHASER
No. 1
Check (√) one TYPE
OF CONTAINER
Glass bottle
Check (√) one TYPE OF PACK
WEIGHT (SIZE)
SHOWN ON LABEL
Write in the amount of
weight per item or
container and tell us the
kind of weight, such as
pounds, and or ounces,
pints, quarts, etc.
THROW
AWAY
$
DESCRIBE SPECIAL
PRICE OR OFFER
CHECK (√) ONE
Kind of Blade
RETURN
Write in all of the words which together
describe the complete brand name.
Write in
number of
packages
or other
containers
(or company
if purchased
from door-to-door
or home delivery
salesperson)
8 PACK
CARBONATED
SOFT DRINKS
NAME OF STORE
Write in:
Number of
blades or
cartridges
per package or
number of
shaving edges
per continuous
band
CARBONATED SOFT DRINKS DIARY TIPS: First of all, enter in the diary only those purchases which are brought home from the store. This includes
items which may later be taken on a picnic or to the beach as well as items which are consumed at home. • Do not enter items which are consumed
at the place of purchase. Usually, these include vending machine purchases, restaurants, refreshment stands, ball games, etc. • When you make a
purchase entry, fill in the answers under every column all the way across the page. • Be sure to enter purchases brought home by all family members.
6 PACK
as shown on label
KIND OF STORE
Check (√) or write in
4 PACK
BRAND NAME
WHERE DID YOU PURCHASE?
GROCERY
OR FOOD
Date
of
month
purchase
was
made
HOW MUCH DID YOU PAY?
HOW
Was this a special price or offer?
MANY
TOTAL
of each
kind did CASH PAID IF
If YES, describe kind, such as
NO,
Don’t
store sale, cents off coupon,
you buy?
include Check
gift on or in package, etc.
here
taxes
How much savings in total?
√
2 PACK
NOTE: Extra space page 1. Free Gifts and Samples, inside front cover.
Write in
DOUBLE EDGE
Other–
Describe
SINGLES
Write in all of the words which together
describe the complete brand name.
(or company
if purchased
from door-to-door
or home delivery
salesperson)
KIT INCLUDING BLADES
Including free blades with
purchase of razor or
travel and shaving kits.
BLADES ONLY
RAZOR BLADES
OF STORE
RAZOR INCLUDING
BLADES
as shown on label
CHECK (√) ONE
Type of Blade
CHECK (√) ONE
NAME
DRUGS
Date of
month
purchase
was
made
KIND OF STORE
Check (√) or write in
DISCOUNT
BRAND NAME
WHERE DID YOU PURCHASE?
Was this a special price or merchandise offer?
HOW
MANY
TOTAL
If YES, describe kind, such as
of each CASH PAID
• Price-Off marked on package
• Store Coupon from newspaper
kind did
Don’t
• Other Coupon from newspaper
IF
you buy? include
• Coupon from magazine
NO,
taxes
or received in mail
Check
on or in package
Write in
here •• Gift
Special refund offer such
number
√
as cash refund or merof
Total
chandise offer
packages
$ ¢
• Store sales • Other deals
¢ off
or other
if any
containers
GROCERY OR FOOD
HOW MUCH DID YOU PAY?
Write in
FIGURE 6-1
Illustrations of recording forms used in mail diary panels.
CONSUMER PURCHASE PANELS
129
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
Advantages of Consumer Panels
The data from a panel can be analyzed as a series of snapshots—providing information on aggregate sales activity, brand shares, and shifts in buyer characteristics and types of retail outlets from
one month to the next. However, just as a motion picture is more revealing than a snapshot, it
is the ability to measure changes in the behavior of individuals that is the real advantage of a panel.
Knowledge of the sequence of purchases makes it possible to analyze:
■ Heavy buyers and their associated characteristics
■ Brand-switching rates and the extent of loyal buying (Evidence of stable purchase activity in
the aggregate usually masks a great deal of individual movement.)
■ Cumulative market penetration and repeat purchase rates for new products (The success of
new products depends jointly on the proportion who have tried them once and then purchased them a second, third, or fourth time.)
A continuous purchase panel is an excellent vehicle for conducting quasiexperiments. A change
in price, advertising copy, or package can be implemented in one panel region and the results compared with those of other panel regions where no change was made. Also, because of the lengthy
relationship with members of continuous panels, there is much more opportunity to collect classification and attitudinal information to help understand the observed changes in behavior.
In comparison with interview methods, although not with audits, the continuous purchase
panel has the advantage of accuracy. Several studies have found that interview respondents will
exaggerate their rate of purchasing (an effect that is most pronounced for infrequently purchased
products) and dramatically oversimplify brand-switching behavior. Apparently, survey respondents tend to equate their most recent brand buying behavior with their “normal” behavior—
whether or not this is accurate.
Limitations of Consumer Panels
130
The limitations all relate to the vulnerability of panels to various biases. The first problem
encountered is selection bias, because of the high rates of refusal and the resulting lack of representativeness. It is estimated that panel recruitment rates may vary from as low as 10 to 15
percent when the initial contact is made by mail in the United States, to 50 percent or more for
personal contacts made on behalf of panels in Great Britain.
A related problem is mortality effect, which may mean a dropout rate in excess of 20 percent
a year. Some of this is unavoidable, because it is attributable to moves and illness. To reduce
both the refusal and mortality rates, all panels offer some incentive for continuing participation;
these include direct money payments and stamp schemes in exchange for gifts.
There is little doubt that those who refuse or drop out differ from those who participate and
remain. In particular, those who are not interested in the topic are the most likely to drop out.
These losses are replaced by new members with similar characteristics. This amounts to a matching procedure, but of course it is not possible to match on all important characteristics. In Chapter 9 we will examine the problem of refusals, which afflicts all forms of survey work. As a
consequence, however, all panels underrepresent the very rich, the very poor, and the transitory.
Panels also are subject to a variety of testing effects. There is a definite tendency for new panel
members to report unusual levels of purchasing because of the novelty of the reporting responsibility. This effect is so pronounced that the first month’s results usually are discarded. Surprisingly,
there is little evidence to suggest that there is any long-run conditioning behavior that would lead
to great brand loyalty or price consciousness that would produce systematically biased data.
Finally, it should be kept in mind that usually one person does the recording of purchases.
Whenever a product, such as cigarettes or toothpaste, is purchased by several members of the
household, there is a good chance that some purchases will be missed. These products are also
SCANNER SERVICES
AND
SINGLE-SOURCE SYSTEMS
troublesome to analyze, for what appears to be brand switching simply may be the purchasing
of different brands for (or by) different members of the household.
SCANNER SERVICES AND SINGLE-SOURCE
SYSTEMS
It is no understatement to say that standardized sources of marketing data for consumer goods
have been revolutionized by the universal product code (UPC) scanner. By the end of 1992, 80
percent of the commodity value of groceries, 85 percent of the commodity value of mass merchandises, and 40 percent of the commodity value for drugs came through scanner-equipped
retail stores.2
Within markets equipped with scanner checkouts, purchases are recorded by passing them
over a laser scanner, which automatically reads the bar-coded description (the universal product
code) printed on each package. This in turn activates the cash register, which relates the product
code to its current price—held in computer memory—and calculates the amount due. All the pertinent data on the purchase then are stored in the computer and can be accessed instantly for analysis. In addition to price, the memory stores information on coupon use, so marketers can quickly
measure consumer response to using coupons across product categories. Information about shelf
space, end-of-aisle displays, use of cooperative advertising, and the like can be merged with scanner data in order to measure the impact on sales, item movement, and net contribution of each
marketing effort. Advanced applications are currently available that convert pictures taken from
mobile phone cameras into scan-quality files that can be used as sources of marketing information.
As mentioned earlier, the impact of scanners is being felt in the conduct of retail audits,
where in-person audits will soon be obsolete, and in the growth of single-source services that
track the behavior of individual households, from the TV set to the checkout counter.
Scanner-Based Audit Services
The most immediate benefits of scanner data are a high degree of accuracy, time savings, and
the ability to study very short time periods of sales activity. To appreciate these benefits, consider the introduction of a new food product into a test market monitored by a scanner-based
audit service.
According to the standard bimonthly reports from the store audit service, there was steady
progress in building the market share. During the second bimonthly period (see graph on this
page), share more than doubled, as retailer advertising and coupon promotions were stepped up.
16.5
Sales share (%)
15.7
6.7
8
16
End of week
24
131
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
27.6
Sales share (%)
23.0
22.0 21.8
18.4
14.9
12.1
12.4 13.1
10.0
6.5
5.1
5.0
2
3
4
5
18.8
17.1
11.3
6.5
4.8
5.1
8.6
1
17.8
6.4
5.8
3.4
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Week
FIGURE 6-2
Weekly results from scanner service.
The results from the third period normally would not be available for six to eight weeks. With
scanning data, however, weekly reports on this product were available within two to three weeks
of the period’s end. These weekly reports were far more revealing, as we see from the data shown
in Figure 6-2 for the same time period following the launch. The first 11 weeks followed a fairly
typical new-brand cycle, with share by week 10 only half of the initial peak. Back-to-back price
and coupon promotion in weeks 12 and 13 boosted shares to twice the introductory level. Shares
then declined, until a further promotion in week 18. Fortunately, postpromotion shares were always
higher than prepromotion shares through week 22. The sharp decline in week 23 was traced to a
shortage of the most popular size of package, which was rectified during the next week.
Three scanner-based audit services provide nationally, or locally, representative results
within two weeks of the end of the reporting period. Each service provides full detail on each
universal product code in a product category: product description, size, price, unit movement
and unit share, and dollar sales and share, as well as availability in stores, shelf-space allocation, and usage of coupons. These data can be made available in virtually any combination
of stores, to look at sales by chain, geographic area, or even individual store.
Nielsen’s SCANTRACK now appears to be the largest service, with access to over 3,000
scanner-equipped stores. Both SCANTRACK and Nielsen Food Index (NFI) reports can be
accessed directly by a manager with a personal computer and a modem, who can then tap
directly into the Nielsen online system. For categories such as candy, snacks, and tobacco, where
a large proportion of sales are made in stores without scanners, the supermarket scanner data
can be combined with in-person audit data. A.C. Nielsen tracks consumer purchasing in over
150,000 retail outlets in more than 65 countries to provide clients with the most accurate, comprehensive, and timely information about the dynamics of the marketplace.
132
Applying Scanner Data. Scanner data are used mainly to study the behavior of the consumer
when different elements of the marketing mix are varied. They are also used to study and forecast
the sales of a new product. For example, in 1991 Frito Lay, Pepsico’s $4.2-billion-a-year snack-food
division, tested its new multigrain snack called Sun Chips for 10 months in the Minneapolis area.
The company experimented with 50 ridges and a salty, nutty flavor, and introduced the product
on supermarket shelves. Using scanner data, the company discovered how many shoppers actually took home their first bag of Sun Chips in a given week and who went back a second and third
time. This “depth of repeat” gives the company a much clearer sense of the product’s potential.3
Even retailers are using scanner data. They not only purchase scanner data, they also conduct their own tests.4 They conduct a number of experiments to analyze historical demand at
various pricing levels and determine the prices at which they can maximize their total contribution. One retail grocer offered Minute Rice at five different prices over a period of 16 weeks,
and the sales results are graphically shown here:5
SCANNER SERVICES
AND
SINGLE-SOURCE SYSTEMS
$1.29
Retail price
1.19
1.09
.99
.89
20
40
60
Average weekly sales of boxes of Minute Rice
80
100
With costs of $0.69 per box, the store could see that profit was maximized at a price of $1.19.
Lower prices did not generate sufficient incremental volume to overcome the drop in per-unit
profit. This kind of analysis can now be done routinely for thousands of retail store items. This
helps retailers control their operations more closely, and some feel that it will significantly
enhance their power over manufacturers. Retailers with strong “store brands” have been able to
capture market share from national brands by better understanding their customers’ preferences
and price sensitivity.
Safeway is an example of a retailer that does a lot of scanner-based research. Safeway used
scanner data to test alternative placement of products within a store. Results showed, for example, that foil-packaged sauce mixes should not be displayed together, but should be spread around
the store according to their contents (spaghetti sauce mix near bottled spaghetti sauce, gravy mix
near canned gravy, etc.). In 1990, Safeway created Safeway Scanner Marketing Research Services
(SSMRS). Its first product was StoreLab. Clients can test the effectiveness of off-shelf displays,
shelf extenders, in-store signs, new package designs or sizes, and consumer bonus packages.6
A new technology called RFID may replace the bar codes. RFID utilizes a tiny silicon chip
to store information; a small transmitter would then send this information to a scanner. RFID
offers more benefits than a UPC, such as:
■ The ability to store more information than an 11-digit bar code
■ The ability to change the information on the tag
■ The ability to transmit all the information on the chip to a scanner without a clear line of sight
RFID tags could make grocery shopping much easier. To check out, all you would have to do
is roll your loaded shopping cart past an RFID scanner. In less than a second, it would scan the
information on every item in your cart. Shoppers might be able to bag the items as they are selected
and walk past the RFID scanner with the items already bagged. Although instant checkout sounds
like a technological benefit every shopper would appreciate right away, RFID tags on groceries are
still years away. The most significant problem is the high cost of the tiny tag components; the price
of a few pennies per tag means that RFID is still impractical for grocery-product manufacturers,
which operate on low profit margins. So while the supermarket industry led the way in the adoption of bar codes, other industries will lead the way to the adoption of RFID tags. Supermarket
shoppers can be certain that the cost of the RFID tags will steadily decrease and will eventually
be less than a penny per tag. Shoppers can also be certain that supermarkets will use any technology that reduces the labor cost required to check out a customer.7 Marketing Research in Action
6-1 illustrates the type of tracking system with bar codes. Bar codes have also facilitated the integration of physical and online transactions of products. Marketing Research in Action 6-2 shows
how use of unique electronic identity (web code) helps producers, retailers, and consumers.
133
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
MARKETING RESEARCH IN ACTION 6-1
Bar Codes Monitor Effectiveness of Ads
David Kirwan of Broadwell Marketing Group has created a bar code tracking system to help clients
determine which types of ads, special offers, and media mix yield the most benefit. Broadwell was
founded seven years ago in a converted two-car garage and has now evolved into a full-fledged marketing and advertising services company that generated $65 million in revenues in 1998. Manufacturers
have used bar code coupons to gather information about consumer buying habits at supermarkets but
Broadwell has adapted it further to get immediate feedback for restaurants and other stores. This is
called intelligent marketing. Pizza Hut has collected $24 million in yearly savings thanks to intelligent
marketing. Since applying this technique in 1994, Pizza Hut has cut its coupon and direct-marketing
costs by 40 percent while nearly doubling its redemption rate. They used the data collected to consolidate different regional marketing departments into one, leveraging its printing and media buying power,
which cuts down promotional pitches from 650 to less than 20. Broadwell clients include about two
dozen major restaurants and retail chains.
For a closer look on this article go to http://search.chron.com/chronicle/advancedSearch.do.
SOURCE: Cheryl Hall, Houston Chronicle, Sunday, March 7, 1999, p. 10D.
MARKETING RESEARCH IN ACTION 6-2
Bar Codes and the Internet
With bar codes, companies can unify the online and physical worlds. Bar codes, or any other machinereadable encoded labels, allow physical objects to send data and instructions to networks. By combining bar codes with the Internet, consumer goods companies can stamp every product with a unique
electronic identity that can be read wirelessly by new scanning gadgets. Stores can take inventory faster,
and more accurately, since shelves equipped with wireless receivers will instantly know when a particular kind of merchandise is gone. With the use of supply chain software and store security system,
stores can track a product’s location by their bar code labels, and stolen items will be identifiable anywhere. Consumers will be able to execute transactions on the Web, and they can scan special “Web
codes” in advertisements or on products.
SOURCE: David Orenstein, “Raising the Bar,” Business 2.0, August 2000.
Single-Source Systems
134
Single-source systems are usually set up in reasonably self-contained communities, with their
own newspaper and cable TV, and are roughly representative of the country’s demographics.8
After recruiting a test panel of community households, with small payments or coupons as
inducements, the researcher monitors each home’s TV viewing habits and quizzes household
members periodically on what newspapers, magazines and web sites they regularly read. This
provides detailed records of exposure to programming and specific commercials.
Each panel member presents an identification card at a scanner-equipped grocery store each
time a purchase is made. This card alerts the checkout terminal to send an item-by-item record
of those purchases to a computer that will be included with all other panel members. Then
researchers can relate the details of a household’s purchase of each product to previously collected classification information about the household and any promotional stimuli to which they
were exposed.
These panel households can also be individually targeted for newspaper advertising, so a marketer can experiment with different combinations of advertising copy and/or exposure, discount
coupons, and in-store price discounts and promotions. The effects of these different programs can
be monitored unobtrusively in the supermarket, and each panel member’s purchase compared with
what he had purchased before the test. To control for competitive activity, the service also tracks the
amount of feature, display, and couponing activity in each supermarket.
SCANNER SERVICES
UPC’s for each
grocery item are
scanned at
checkout.
Information is sent from store
to chain and on to IRI via
telecommunication systems.
Household
panel members
present an
identification
card at checkout, which
identifies and assigns items
purchased to that household.
Coupons are collected and
matched to the appropriate
UPC. Information is
electronically communicated
to IRI computers.
IRI field
personnel
visually survey
stores and all
print media to record retailer’s
merchandising efforts, displays,
and ad features. Field personnel
also survey retail stores for a
variety of custom applications,
e.g., average number of units
per display, space allocated to
specific sections, and number
of facings. Results are
electronically communicated
to IRI computers.
AND
SINGLE-SOURCE SYSTEMS
Household
panel members
are selected for
television
monitoring and equipped with
meters, which automatically
record the set’s status every
five seconds. Information is
relayed back to IRI’s computers.
Completed databases are converted to
the required client format, transferred
to appropriate recording mediums...
hard copy, mag
tape, pc diskette,
etc..., and sent
to the subscriber.
Information is received at IRI and processed through the Neural
Network (Artificial Intelligence) Quality Control System. The
system approves the over 35 million records obtained weekly for
further data processing and identifies records that require
further verification.
FIGURE 6-3
The process of scanner data collection by IRI.
SOURCE: Information Resources, Inc.
Television advertising exposure also can be controlled through the cable provider. For example, it is possible to transmit a Duncan Hines cake mix commercial only to Betty Crocker
customers to find out if they can be induced to switch. The process of collecting data and providing it to manufacturers and retailers is shown in Figure 6-3.
A typical application of single-source systems is an experiment by the Campbell Soup Company. Using an index of 100 for the average household’s V8 consumption, Campbell found that
demographically similar TV audiences were consuming vastly different quantities of V8. In early
1987, for example, General Hospital audiences had a below-average 80 index, while Guiding Light
audiences had an above-average 120 index. The results were surprising because General Hospital
had a slightly higher representation of 25- to 54-year-old women, who were known from other
research to be heavier buyers of V8. With this sort of information at hand, it becomes possible
to rearrange media schedules to raise the average index.9
The possibilities for research are limited only by the ingenuity of the researchers.10 Suppose
the research purpose is to decide whether a product sample should be an alternative or an accompaniment to TV commercials for a child-oriented cereal. The household test panel could be
divided into four matched groups and given different combinations of samples and advertising.
Sales results could then be monitored to determine whether the group that was also exposed
to advertising would be more brand-loyal in subsequent weeks.
The advantages of single-source data for this kind of market response study—compared with
conventional market tests—are (1) availability of extensive pretest records, (2) immediate availability of test results, and (3) ability to compare the purchases of households receiving a specific
135
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
ad during the test with their own purchases prior to the test as well as with purchases of those
who were not exposed to the new product ad.
The disadvantage is that single-source systems can track purchases only at grocery stores
and drugstores that are equipped with scanners. Furthermore, the test services do not know
whether viewers actually are watching a television commercial when it is on, or whether they
leave the room during the commercial break. Test participants who are paid volunteers might
also unconsciously bias the results, because they know they are sending a message to advertisers. An unresolved question is whether tests conducted in small, self-contained markets can
predict nationwide results. Finally, there are significant differences in results among competing
single-source services that raise questions about the quality of the findings.
Comparing Single-Source Services. One of the most advanced services is BehaviorScan, by
Information Resources, Inc. (IRI).11 It has 3,000 households wired in each of various minimarkets across the United States, from Pittsfield, Massachusetts, to Visalia, California. BehaviorScan,
an electronic test marketing service, allows marketers to test new products or new marketing
programs. BehaviorScan targets TV commercials in a controlled environment to specific households and tracks the effectiveness of those ads in six smaller markets. It pioneered the service
and has had the most experience in conducting tests. IRI’s system uses the identification card
described earlier, so that details of transactions can be recorded, by household, upon presentation of the card to cashiers at participating stores. IRI also has another product called InfoScan.
InfoScan is a syndicated market tracking service that provides weekly sales, price, and storecondition information on products sold in a sample of 12,000 grocery and 7,800 drug stores.
The InfoScan service also includes a national sample of thousands of households whose purchases are recorded by checkout or hand-held scanners. BehaviorScan provides data about
different markets, whereas InfoScan provides data on different products. Retail sales, detailed
consumer purchasing information, and promotional activity are monitored and evaluated for
all UPC products.
For several years A.C. Nielsen had been testing a single-source service—Consumer Panel
Services—and had 15,000 homes by 1989. In 1999, Nielsen had increased its panel size to 125,000
households. Nielsen, along with another new entrant, Scan America (a joint venture of Control
Data and Selling Areas Marketing, Inc., a subsidiary of Time, Inc.) offers a very different technology. Unlike BehaviorScan participants, Nielsen families record their purchases at home by
passing a penlike wand over goods bearing bar codes, keying in other information about coupons
at the same time. The information is transmitted daily through a device attached to participants’
television sets. Because Nielsen’s system is not dependent on retailers who cooperate, it allows
tracking of purchases from a wider range of stores, but it has been criticized for requiring that
participants be actively involved in entering the data. Another important difference is that
Nielsen can send test commercials over the air, so members of the sample do not have to be
cable subscribers.12
Retail Measurement Services provides product movements data through food, drug, and
other retail outlets in 79 countries. Flagship service is U.S.-based Scan-Track, which provides
weekly data on packaged-goods sales, market shares, and retail prices from UPC scannerequipped supermarkets. Another service, Precision, tracks health and beauty aids product sales
through drug and mass-merchandiser stores. In other countries, either scanning or audit-based
methods are used to collect retail data.
136
Expert Systems Based on Single-Source Services. The rapid increase in the use of scanner technology has brought a peculiar problem to the marketing analyst. Although about a decade ago
researchers were struggling to obtain data to conduct research, they now are faced with exactly
the opposite problem. Users of scanner data have been overwhelmed with massive amounts of
data. Therefore, suppliers of scanner data have created expert systems to give the data more
utility for the managers. These expert systems are designed to cut the mass of scanner data down
to actionable pieces of information. There are many expert systems that give solutions to specific
MEDIA-RELATED STANDARDIZED SOURCES
problems. Both Aldata and Nielsen have developed a number of these expert systems to make
the decision process easier. A partial list is provided below.
Aldata13
■ Apollo Space Management Software provides suggestions for optimizing shelf allocations
for each item in the section of a retail store. It analyzes scanner data from the InfoScan
database to review the amount of shelf space, price, and profit components of product
category shelf sets, such as dishwashing soaps or cereals. Apollo also can produce photoquality schematics using its library of 120,000 product images and dimensions. Thus, the
retailer gets a visual picture of what the shelf reallocation will look like.
Nielsen14
■ Promotion Simulator is an easy-to-use software application that automatically simulates
future promotion strategies. It uses historical Nielsen data to calculate the effect of the promotional plans on manufacturer and retailer profitability, given the price points and retail conditions in the plan.
■ Spotlight is an expert system that runs through the database, calculating volume and share
changes, and searching for the key merchandising and competitive factors that influence those
factors. Then it summarizes the findings in a presentation-quality report.
■ Sales Advisor automatically highlights need-to-know information and enables the sales force
to add personal insights to presentation-quality output. Sales Advisor produces finished summaries and presentations of sales and marketing data.
MEDIA-RELATED STANDARDIZED SOURCES
Another area in which there is a great deal of commercial information available for marketers
relates to advertising and media. A number of services have evolved to measure consumer exposure to the various media and advertisements. For example, Nielsen Media Research unveiled
its much-anticipated Nielsen/Net Ratings Internet Measurement Service on March 22, 1999.
According to a Nielsen spokesman, the NetRatings service uses data from 9,000 users on a realtime basis and will track movement on both websites and so-called banners, or Web-page ads
that users click on to learn more information. Nielsen took on the help of Media Metrix, which
has been the main company measuring consumer use of websites.15 In addition to Nielsen,
several media-related services are described below.
Nielsen Television Index
The Nielsen Television Index (NTI) is probably the best known of all the commercial services
available in this category. As a system for estimating national television audiences, NTI produces
a “rating” and corresponding share estimate. A rating is the percent of all households that have
at least one television set tuned to a program for at least 6 of every 15 minutes that the program
is telecast. Share is the percent of households that have a television set that is tuned to a specific
program at a specific time.
From 1987 onward, Nielsen has been using “people meters” instead of the traditional diary
to obtain this information. The people meter, which is attached to the television set, continuously
monitors and records television viewing in terms of when the set was turned on, what channels
were viewed, how long the channels were tuned in, and who in the household was watching that
channel.16 The data are stored in the people meter and later are transmitted via telephone lines
to a central processing facility.
137
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
Nielsen National Television Index (NTI) is based on a national 5,000-household sample
equipped with people meters, and reporting TV, cable, and home video viewing. At the market
level, Nielsen Media Research (NMR) provides Nielsen Station Index (NSI), which uses set
meters to measure TV audiences through weekly mail diary surveys in 211 local markets among
100,000-plus households. Nielsen Homevideo Index (NHI) measures cable and VCR audiences
using a combination of NTI, NSI, and custom diaries. Data on syndicated programs in the
Nielsen Syndication Service (NSS) come from the same sources. Two services, through people
meters, measure Hispanic audiences for Spanish-language media across the United States (800
households) and in the Los Angeles market (200 households).
The Nielsen NTI-NAD data are used by media planners to analyze alternative network programs. Next, the cost efficiency of each program is calculated. Cost efficiency represents a television program’s ability to deliver the largest target audience at the smallest cost. A cost per
thousand (CPM) calculation is computed as
Cost of a commercial
⫽ CPM
Number of target audience delivered
Since 1994, Statistical Research Inc. (SRI)—a company known primarily for measuring national
network ratio audiences—has made a priority of its pilot program for a new TV measurement
system called SMART (Systems for Measurement and Reporting Television). SMART is known
for its user-friendliness and the ability to count more viewers than Nielsen’s People Meter. Media
industry observers believe SRI could have what it takes to topple Nielsen.
SRI’s SMART grew out of the broadcast industry’s dissatisfaction with Nielsen Media
Research. Critics have complained that Nielsen does not measure enough viewers, charges too
much, and takes too long to provide subscribers with data. On the other hand, the SMART setup
consists of meters with sensors that can pick up signals from the air. The meters require no wire
connection and the meters attach to television sets with a Velcro strip.
Despite SRI’s potential, the company still faces an uphill battle against Nielsen. According
to president and co-founder of SRI Gale Metzger, SMART was designed to respond to the industry’s needs, and whether or not his company will beat out Nielsen depends on the industry.17
Major differences between the two systems are given in the chart below:
SMART
People Meter
Mechanics
No wire connection—sensors pluck signals from
the air.
Wired directly to TV and
VCR tuners.
Research
Methods
Data are retrieved by reading embedded codes
called Universal Television Program Codes
(UTPC) in some program signals; a decoder
attached to the back of the TV will be able to
record the UTPC identifying the program used.
Telephone connections are
used to return data.
Method of Data
Collection
User logs in and out before and after watching
TV with remote control device.
Reputation as:
Media measurement business serving the radio
industry.
User punches numerical
code into a data-entry
device.
Arbitron Diary Panel
138
Arbitron gathers information that describes consumers’ demographic, socioeconomic, and
lifestyle characteristics, as well as purchase intentions in 260 markets. Arbitron provides
three qualitative services tailored to fit specific market size and marketing requirements:
Scarborough caters to large markets, RetailDirect® is available in medium markets, and the
Qualitative Diary Service is offered in smaller markets ranked 100⫹. Each service profiles a
MEDIA-RELATED STANDARDIZED SOURCES
market, the consumers, and the media choices in terms of key characteristics. The three services cover the major retail and media-usage categories in almost any area.
There has been a lot of recent controversy over Nielsen television ratings. The networks claim
that since Nielsen became a monopoly—all the other rivals dropped out—service has been unprofessional. The networks also have problems with Nielsen’s sampling methodology. The networks
charged that Nielsen bases its ratings on too few diaries, making its data less representative of
what viewers watch. In response to this, Nielsen decided to send out 10 percent more paper
diaries for viewers to note what they watch. By 1998 the number of diaries increased by 50 percent,
which Nielsen claims is the largest increase in sample size in the history of TV audience measurement.
Arbitron, a subsidiary of Control Data, maintains both national and regional radio and TV
panels. The panel members are chosen by randomly generated telephone numbers, to ensure that
households with unlisted numbers are reached. Those household members who agree to participate when called are sent diaries in which they are asked to record their radio listening behavior over a short duration. Most radio markets are rated only once or twice a year; however, some
larger ones are rated four times a year. The TV diary panel is supplemented with a sample of
households that have agreed to attach an electronic meter to their television sets. Arbitron
produces custom reports for clients. Typically, these are based on an interactive computer-based
system called Arbitron Information on Demand (AID).
Starch Scores
The previous two sources of marketing data involve radio and television ratings, whereas Starch
scores cover print media. The Starch Readership Service measures the readership of advertisements in magazines and newspapers.18 Some 75,000 advertisements in 1,000 issues of consumer
and farm magazines, business publications, and newspapers are assessed each year, using 100,000
personal interviews.
The Starch surveys employ the recognition method to assess a particular ad’s effectiveness.
Four degrees of reading are recorded:
1. Nonreader: A person who does not remember having seen the advertisement in the issue.
2. Noted: A person who remembers seeing the advertisement in the issue.
3. Associated: A person who not only “noted” the advertisement, but who also saw or read some
part of it that clearly indicated the brand or advertiser.
4. Read Most: A person who read 50 percent or more of the written material in the ad.
Because newspaper and magazine space cost data are also available, a “readers per dollar”
variable can be calculated. The final summary report from Starch shows each ad’s (one-half page
or larger) overall readership percentages, readers per dollar, and rank when grouped by product category.
Multimedia Services
Simmons Media/Marketing Services uses a national probability sample of some 19,000 respondents and serves as a comprehensive data source, allowing the cross-referencing of product usage
and media exposure. Four different interviews are conducted with each respondent, so that magazine, television, newspaper, and radio can all be covered by the Simmons Service. Information
is reported for total adults and also for males and females separately.
Media-mark Research also makes available information on exposure to various media and
on household consumption of various products and services. Its annual survey of 20,000 adult
respondents covers more than 250 magazines, newspapers, radio stations, and television channels, and over 450 products and services.
139
CHAPTER 6 STANDARDIZED SOURCES
OF
MARKETING DATA
APPLICATIONS OF STANDARDIZED
SOURCES OF DATA
Standardized data sources are used widely in a number of marketing research problems. We have
already discussed the various sources of standardized data. In this section we will look at some
of the applications of standardized data. Table 6-2 gives a comprehensive framework of the types
of sources used for different applications.
Measuring Product Sales and Market Share
A critical need in today’s increasingly competitive environment is for firms to have an accurate
assessment of their performance. Marketing performance typically is measured by the company’s
sales and market share. Firms can track their own sales through analyses of their sales invoices.
An alternative source that can be used is the online bibliographic data sources. The measurement of sales to final customers historically has been done by the use of diary panels and retail
store audits. Scanner data is now being widely used for this purpose.19
Measuring Advertisement Exposure and Effectiveness
With media costs rising exponentially, and the options for advertising increasing dramatically
with the advent of cable television, measurement of advertising exposure and effectiveness and
program ratings has become critical. Advertis-ing exposure and effectiveness in the print media
are tested by Starch scores20 and in the audio-visual media by Nielsen Television Services and
Arbitron. Ratings of the various network and cable programs can be found in the Nielsen Television Index, and those of the radio programs in the Arbitron panels. Multimedia services such
as the Simmons Media/Marketing Services and Media-mark Research conduct research on overall media exposure and effectiveness. Scanner data also have been used extensively for modeling ad exposure and measuring ad effectiveness.21
TABLE 6-2
Applications of Standardized
Data Sources
Measuring Promotion
Effectiveness
Measuring Ad Exposure
and Effectiveness
1. Scanner data
1. Starch scores
2. Diary panels
2. NTI
3. Arbitron
4. Multimedia services
Measuring Product Sales
and Market Share
1. Diary panels
1. Scanner data
2. Retail audits
2. Starch scores
3. Scanner data
3. Diary panels
4. Internal records
4. Internal records
5. NAICS
140
Estimation and Evaluation
of Models
QUESTIONS
AND
PROBLEMS
Measuring Promotion Effectiveness
Only since the advent of scanner data have marketers begun to fathom the power of promotion
in increasing product sales and market share. This is evident from the fact that the number of
coupons, samples, displays, and features at the retail outlets have increased tremendously over
the past few years. Before scanner data was introduced in a big way, promotion effectiveness
was measured by diary panels of the NPD group. Since store-level scanner data and scanner
panels were introduced by IRI and Nielsen, measurement of promotion effectiveness has been
revolutionized. The effect of promotions now can be observed directly through scanner data.22
Estimation and Evaluation of Models
Models are essentially explanations of complex phenomena, expressed by symbols. They can be
mathematical, verbal, graphical, or a figure. Marketing researchers build models to explain the
various marketing phenomena and to draw managerial implications from them. The models
can either be at the aggregate level or at the individual level. Scanner data are used to evaluate
the efficacy of the model and estimate the response of the market to changes in the marketingmix elements.23 North American Industry Classification System (NAICS) codes and diary panels
are used for this purpose.
Summary
Between the generally available secondary sources (which are economical and quickly found but perhaps not
relevant) and primary data (which are designed to be directly relevant but consume considerable time and
money), there are various standardized information services. These exist whenever there are economies of scale
in collecting data for a number of users with similar information needs. The most widely used services are still
retail store audits and consumer panels. With recent developments in information technology, and especially
the widespread adoption of store checkout scanners, the use of these services is quickly shifting to scanner
audits. New capabilities for understanding consumer response by integrating measures of purchase behavior
and communications exposure in single-source data are also changing the conduct of research very rapidly.
The remainder of this book will emphasize the collection and analysis of primary data, for it is here that
the greatest problems and opportunities are found. Knowledge of the methods of primary data collection is
also essential to informed usage of secondary sources.
Questions and Problems
1.
2.
Which of these two, a product audit (such as the
Audits and Surveys National Market Audit) or a
store audit (such as the Nielsen Retail Index),
would be more suitable for the following products:
(a) peanut butter; (b) cameras; (c) engine oil additives; (d) chewing gum? Why?
The manager of a supermarket in Hoboken, New
Jersey, has two local newspapers competing for
weekly ads. The salesperson for newspaper A
claims that although A’s rates are significantly
higher than newspaper B’s, paper A has the best
circulation in the market. Also, readership studies
show that area residents really scrutinize the
paper’s pages. The salesperson for newspaper B
concedes that the publication does not have the
widest circulation, but points to readership studies
3.
4.
showing that its ads are equally effective. “Why
pay twice as much for an ad, when my paper can
do as strong a job for less money?” This problem
has been unresolved for some time. Recently,
the store had nine scanner checkouts installed. The
manager would like your advice on how to use
the store’s scanner sales data to compare the effectiveness of the two newspapers.
What combination of standardized services would
you use to monitor the effects on your existing
cereal brands when a new cereal brand is being
introduced? Why? Would you use the different
services to monitor trial and repeat purchases of
the new cereal product in a test market?
The sales and market shares for the same brand in
the same period are often different when they are
141
5.
6.
measured by a scanner audit rather than a consumer panel. How would you account for the differences? What would you do about them?
Campbell Soup used BehaviorScan data to discover that viewers of two daytime serials had very
good appetites. Search for Tomorrow viewers bought
27 percent more spaghetti sauce than average, but
22 percent less V8 vegetable juice. Viewers of All
My Children bought 10 percent less spaghetti sauce
than average, but purchased 46 percent more V8
than the norm for all viewers. How would you go
about explaining the differences between these two
programs that led to the consumption differences?
What action would you recommend to Campbell
Soup management?
Wash O’Well, a leading detergent company in
Canada, is unable to explain the reason why its
new zeolite-based detergent has not captured the
anticipated market share since its launch in
7.
Minneapolis, Minnesota, in 1992. The company
has decided to focus on point-of-sale activities: for
example, which brands have distribution in
which stores, what shelf location each brand occupies, the kinds of displays in effect, and so forth.
The marketing manager has been instructed to
recommend a suitable source of standardized
data.
a. What are the various data source alternatives
available to Wash O’Well?
b. Discuss the advantages and limitations of each
of them and recommend the data source most
suitable for Wash O’Well’s needs.
a. What are the two types of consumer panels
used for gathering marketing research data?
b. Which type of panel is most frequently used in
the United States?
c. What are the advantages and disadvantages of
using a consumer panel?
End Notes
1. Seymour Sudman and Robert Ferber, Consumer Panels, Chicago: American Marketing Association, 1979.
2. Source: A.C. Nielsen.
3. Susan Caminiti, “What the Scanner Knows About
You,” Fortune, December 3, 1990, pp. 51(2).
4. James Sinkula, “Status of Company Usage of Scanner Based Research,” Journal of the Academy of Marketing Science, Spring 1986, p. 17.
5. This example was adapted from Leonard M. Lodish
and David J. Reibstein, “New Gold Mines and
Mine Fields in Marketing Research,” Harvard Business Review, January–February 1986, pp. 168–182.
6. “Safeway Launches New StoreLab Testing Service,”
Marketing News, January 8, 1990, p. 45.
7. Martin Sloane, “Technology May Alter Supermarket
Checkout,” Houston Chronicle, Wednesday, July 21,
1999, p. 11F.
8. The following articles give a good overview of the
single-source systems: David Curry, “Single Source
Systems: Retail Management Present and Future,”
Journal of Retailing, Spring 1989, pp. 1–20; Melvin
Prince, “Some Uses and Abuses of Single Source
Data for Promotional Decision Making,” Marketing
Research, December 1989, pp. 18–22; “Futuristic
Weaponry,” Advertising Age, June 11, 1990, pp. 5–12.
9. Joanne Lipman, “Single Source Ad Research Heralds
Detailed Look at Household Habits,” Wall Street
Journal, February 16, 1988, p. 35.
10. Examples of recent applications of scanner data
include: Peter S. Fader and Leigh McAllister, “An
Elimination by Aspects Model of Consumer
Response to Promotion Calibrated on UPC Scanner
Data,” Journal of Marketing Research, August 1990, pp.
322–332; Eric Waarts, Martin Carree, and Beren
Wierenga, “Full Information Maximum Likelihood
142
11.
12.
13.
14.
15.
16.
17.
18.
19.
Estimation of Brand Positioning Maps Using Supermarket Scanning Data,” Journal of Marketing Research,
November 1991, pp. 483–490.
The information on BehaviorScan is drawn from a
pamphlet from Information Resources Incorporated.
A good article that compares the two single-sources
suppliers is Leon Winters, “Home Scan vs. Store
Scan Panels: Single Source Options for the 1990s,”
Marketing Research, December 1989, pp. 61–65. Also
see “Now It’s Down to Two Equal Competitors,”
Superbrands 1991: A Supplement to Adweek’s Marketing
Week, p. 28; and “IRI, Nielsen Slug It Out in Scanning Wars,” Marketing News, September 2, 1991, p. 1.
Source: Information Resources Incorporated.
Source: A.C. Nielsen.
“Nielsen on the Net,” Wall Street Journal, Friday,
March 19, 1999.
William R. Dillon, Thomas J. Madden, and Neil H.
Firtle, Marketing Research in a Marketing Environment, Boston: Irwin, 1990; and Soong Roland, “The
Statistical Reliability of People Meter Readings,”
Journal of Advertising Research, February–March
1988, pp. 50–56.
Michelle Wirth-Fellman, “A SMART Move,” Marketing News, September 14, 1998, pp. 1, 7.
Starch Readership Report: Scope, Method and Use,
Mamaroneck, NY: Starch INRA Hooper, undated.
V. Kumar and T. Heath, “A Comparative Study of
Market Share Models Using Disaggregate
Data,” International Journal of Forecasting, July 1990,
pp. 163–174; Francis J. Mulhern and Robert
P. Leone, “Implicit Price Bundling of Retail
Products: A Multiproduct Approach to Maximizing Store Profitability,” Journal of Marketing,
October 1991, pp. 63–76.
20. George M. Zinkhan and Betsy D. Gelb, “What
Starch Scores Predict,” Journal of Advertising
Research, August–September 1986, pp. 45–50.
21. Gerald J. Tellis, “Advertising Exposure, Loyalty,
and Brand Purchase: A Two-State Model of
Choice,” Journal of Marketing Research, May 1988,
pp. 134–144; V. Kumar and R. T. Rust, “Market Segmentation by Visual Inspection,” Journal of Advertising Research, 1989, pp. 23–29.
22. Sunil Gupta, “Impact of Sales Promotions on When,
What, and How Much to Buy,” Journal of Marketing
Research, November 1988, pp. 342–355; V. Kumar and
R. Leone, “Measuring the Effect of Retail Store
Promotions on Brand and Store Substitution,”
Journal of Marketing Research, May 1988, pp. 178–185;
Robert C. Blattberg and Scott A. Neslin, Sales Promotion: Concepts, Methods, and Strategies, Englewood
Cliffs, NJ: Prentice Hall, 1990.
23. V. Kumar, A. Ghosh, and G. Tellis, “A Decomposition of Repeat Buying,” Marketing Letters, 3(4),
1992, pp. 407–417; Peters S. Fader, James M. Lattin,
and John D. C. Little, “Estimating Nonlinear Parameters in the Multinomial Logit Model,” Marketing
Science, Fall 1992, pp. 372–385.
CASE 6-1
Promotion of Rocket Soups
The data for this case were collected from grocery store
scanner installations in one market area, and consist of
Soups provide an interesting case history through
which to examine the impact of sales promotion. Soups
are sold primarily through food stores and frequently
are promoted by both the manufacturer and the retailer.
This case is based on 40 weeks of scanner sales history
in one city. Please answer all of the questions at the end
of the case.
Background
Soups are purchased by over 75 percent of all households. On average, the time between purchases is
around 40 days. Sales promotion, including features,
displays, and price reductions, is extremely important
in this category, with around 55 percent of annual category sales moving with some sort of sales promotion.
Shelf-price reductions, when they are used, average
about 30 percent.
The major competitors in this category are
Brands
Rocket
Stellar
Tasty
Lovely
Happy
Smile
Share
19.7
40.3
13.4
3.5
18.1
5.0
1. Weekly store sales by brand
2. Predominant price charged by brand by store
3. Occurrence of promotional activity by brand by store
The data have been summarized across all stores
in the market on a weekly basis, for a total of 60 weeks.
Table 6-3 is the week-by-week summary.
At the weekly level, Table 6-4 shows the data on
the amount of volume moved by various promotions
at various prices. However, since the stores in the market are of varying sizes, direct evaluation of promotional effects is difficult. Therefore, the sales effects are
normalized (adjusting for store sizes statistically).
Table 6-5 shows the normalized sales volume by promotional type. The commonly used methods for storelevel scanner data are (1) brand volume per $1,000,000
total all commodity volume spending, and (2) brand
volume per 1,000 register checkouts.
During the time of data collection, some extremely
“hot” promotional activity occurred as the result of a
Rocket promotional program. Some retailers decided to
use Rocket soup as a “loss leader” in addition to the
manufacturer promotions, and combined both. These
activities resulted in substantial variations in price, volume, and share of Rocket.
This case was prepared by V. Kumar for the purpose of classroom discussion. For similar cases,
refer to J. C. Totten and M. P. Block, Analyzing
Sales Promotion, Chicago: Dartnell Corp., 1994.
Case Questions
Question 1
According to Table 6-3, weeks 5, 12, 22, 23, 36, and 37
were highly important sales weeks for Rocket. Weeks
2, 8, and 31 were important sales weeks for the compe-
tition. If data on promotional activity are not known, its
presence is usually inferred by abnormally high sales
of the brand or category, a cut in price, or both. Reviewing these weeks, and nearby weeks, discuss the effects
on Rocket and its competitors on:
143
1.
2.
3.
4.
Total category volume
Total category dollar spending
Each other’s volume (Rocket’s and competitors’)
Sales in the weeks following high volume movement
Question 2
Table 6-4 shows overall promotional activity. Sales during week 12 were at an average price near zero. This
typically arises out of a retail promotion of the type:
“Free box of Rocket given with total purchases of
$10.00 or more.” Table 6-5 shows sales levels adjusted
for the percentage of stores in each class, and for their
relative sizes. Discuss any difficulties you see in
assessing the impact of the week 12 promotion on:
Question 3
Using Table 6-5, plot only the nonpromoted sales versus the nonpromoted price. Discuss the probable percent impact on Rocket sales from a 10 percent increase
or decrease in price.
Question 4
Using Table 6-5, plot the relationship between sales on
display only, and display-only price. What similarities
and differences in sales levels and response to price do
you see in comparing the display-only response to the
nonpromotional response?
1. Incremental volume to Rocket
2. Promotional volume borrowed from future sales
CASE 6-2
Kerry Gold Products, Ltd.
In late May of 1996 the research manager for Kerry
Gold Products met with the product manager for margarine to review the company’s first experience with
Nielsen scanner data. A year-long test in a single chain
organization, which began in April 1995, had been
completed recently.
The first purpose of their review was to interpret
the findings. They decided to concentrate on the results
of the first 18 weeks of the tests, which are summarized
in Figure 6-4. The size of the bars represents the weekly
unit sales for Kerry Gold brand and the three competing brands also sold by the chain organization. In
addition to the weekly data there were summary data
CASE 6-3
Paradise Foods1
by Steven H. Star and
Glen L. Urban
Introduction
Bill Horton sat alone in his office late Friday afternoon
anxiously leafing through computer printouts, even
1
Steven H. Star is director of the marketing center at MIT’s Sloan
School of Management. Glen L. Urban is Dai-Ichi Kangyo Bank
Professor of Management and deputy dean of the Sloan School. This
case is reprinted with permission from Harvard Business Review,
September–October 1988.
144
on share of total sales for the first and second halves of
the year. This distinction was important because
Kerry Gold had spent relatively little on promotion in
the second half of the year. Since Nielsen had full
records of all sales, they were also able to examine their
share of sales during weeks when no brands offered
price-reduction promotions. These data are shown in
the two right-hand columns in Table 6-6.
While reviewing these results, the research manager also was wondering whether scanner data would
be useful for other grocery products sold by Kerry
Gold. Many of them were as heavily promoted as margarine. Judgments on the desirability of consumer promotions for these products usually were based on a
combination of store audit data plus periodic controlled
experiments.
though he could recite their contents from memory.
Horton was waiting for his boss, Bob Murphy, to report
back the decision on a subject the marketing committee had been debating for more than four hours. The
issue—whether Paradise Foods should authorize
national rollout of a new product, Sweet Dream, to
complement its established frozen specialty dessert,
LaTreat. Horton was product manager for Sweet
Dream, and Murphy was the group manager responsible for all new products in Paradise’s dessert line.
“I’m glad you’re sitting,” Bob quipped uncomfortably as he entered Bill’s office. “The news isn’t good.
The committee decided not to go ahead.”
“I don’t believe it,” Bill protested, “I started to
worry when the meeting dragged on, but I never
TABLE 6-3
City Sales of Soup by Week (Price is Volume-Weighted Average)
Week
1
Rocket
Volume
306
Rocket
Dollars
Category
Volume
Category
Dollars
Rocket
Price
250
5,984
4,356
0.82
Rocket
Share
Competitive
Volume
5.1
5,678
2
345
384
14,288
9,780
1.11
2.4
13,943
3
873
694
6,201
4,893
0.79
14.1
5,328
4
900
715
5,657
4,447
0.79
15.9
4,757
5
4,404
2,772
7,673
5,453
0.63
57.4
3,269
6
456
388
5,027
4,196
0.85
9.1
4,571
7
444
395
5,570
4,687
0.89
8.0
5,126
8
513
507
11,519
8,305
0.98
4.5
11,006
9
483
441
4,661
3,917
0.91
10.4
4,178
10
615
548
4,803
4,381
0.89
12.8
4,188
11
624
571
5,038
4,632
0.91
12.4
4,414
12
16,113
705
19,633
3,982
0.04
82.1
3,520
13
1,056
811
4,476
3,967
0.76
23.6
3,420
14
381
375
3,863
3,599
0.98
9.9
3,482
15
534
526
5,014
4,677
0.98
10.6
4,480
16
1,821
870
6,487
5,257
0.47
28.1
4,666
17
390
384
3,642
3,350
0.98
10.7
3,252
18
405
398
3,743
3,441
0.98
10.8
3,338
19
1,152
1,012
5,042
4,865
0.87
22.8
3,890
20
348
342
4,733
4,248
0.98
7.4
4,385
21
378
372
3,862
3,448
0.98
9.8
3,484
22
5,319
5,129
8,352
7,839
0.96
63.7
3,033
23
2,280
1,691
7,086
5,311
0.74
32.2
4,806
24
1,194
962
5,931
4,878
0.80
20.1
4,737
25
1,020
837
5,708
4,666
0.82
17.9
4,688
26
342
336
7,780
5,849
0.98
4.4
7,438
27
1,509
1,378
5,385
5,191
0.91
28.0
3,876
28
420
413
7,352
5,533
0.98
5.7
6,932
29
528
520
7,516
5,661
0.98
7.0
6,988
30
942
676
6,416
5,461
0.71
14.7
5,474
31
957
711
15,599
7,578
0.74
6.1
14,642
32
1,014
738
4,937
4,294
0.72
20.5
3,923
33
1,059
990
5,664
5,213
0.93
18.7
4,605
34
351
385
9,045
7,302
1.09
3.9
8,694
35
384
418
5,435
4,932
1.09
7.1
5,051
36
4,122
3,084
8,672
7,699
0.74
47.5
4,550
37
9,108
4,466
13,165
8,694
0.49
69.2
4,057
38
498
542
5,643
5,103
1.09
8.8
5,145
39
459
500
5,555
5,150
1.09
8.3
5,096
40
486
529
4,485
4,285
1.09
10.8
3,999
145
TABLE 6-4
City Sales of Soup by Week (Price is Volume-Weighted Average Sales by
Promotional Type)
Week
Volume
with No
Promotion
Price
with No
Promotion
Volume
with
Display
Only
Price
with
Display
Only
Volume
with
Feature
Only
Price
with
Feature
Only
Volume
with
Feature
and
Display
Price
with
Feature
and
Display
1
306
0.82
0
.
0
.
0
.
2
345
1.11
0
.
0
.
0
.
3
477
0.75
396
0.85
0
.
0
.
4
420
0.75
480
.
5
222
0.69
0
6
399
0.83
7
375
8
9
0.84
0
.
0
.
0
.
4,182
57
0.99
0
.
0
.
0.87
69
0.99
0
.
0
.
471
0.98
42
1.07
0
.
0
.
405
0.90
78
0.99
0
.
0
.
10
486
0.87
129
0.99
0
.
0
.
11
537
0.90
87
0.99
0
.
0
.
12
486
0.79
102
13
384
0.75
0
0.63
0.85
0
.
15,525
0.02
.
0
.
672
0.78
14
381
0.99
0
.
0
.
0
.
15
534
0.99
0
.
0
.
0
.
16
267
0.87
153
0.98
0
.
1,401
17
390
0.99
0
.
0
.
0
.
.
0
.
0
.
18
405
0.98
0
19
315
0.81
672
0.89
60
0.98
105
0.35
0.98
20
348
0.99
0
.
0
.
0
.
21
378
0.98
0
.
0
.
0
.
22
201
0
.
0
.
5,118
23
0
24
25
0.69
0.98
.
702
0.86
1,578
819
0.86
375
0.69
366
0.74
654
0.86
26
252
0.98
90
27
117
0.99
288
28
327
0.98
29
309
0.98
30
588
0.74
31
450
0.73
32
333
0.69
33
555
0.87
34
303
1.10
48
1.12
0
.
0
.
35
384
1.09
0
.
0
.
0
.
36
132
0.89
0
.
0
.
3,990
37
0
0.74
0
.
6,699
38
498
1.09
0
.
0
.
0
.
39
459
1.09
0
.
0
.
0
.
40
486
1.09
0
.
0
.
0
.
146
.
0.69
0
.
0
.
0
.
0
.
0
.
0.99
0
.
0
.
0.99
585
93
0.99
0
.
0
.
219
0.99
0
.
0
.
354
0.69
0
.
0
.
507
0.75
0
.
0
.
288
0.69
195
504
1.01
0
2,409
0.89
0.79
.
519
198
0
0.88
0.79
.
0.74
0.40
TABLE 6-5
City Sales of Soup by Week (Price is Volume-Weighted
Average Normalized Volume Sales by Promotional Type)
Week
Normalized
Volume
with No
Promotion
Price with
No
Promotion
Normal
Volume
with
Display
Only
Price
with
Display
Only
Normal
Volume
with
Feature
Only
Price
with
Feature
Only
Normal
Volume
with
Feature
and Display
Price
with
Feature
and
Display
1
0.21
0.82
0.00
.
0.00
.
0.00
.
2
0.20
1.11
0.00
.
0.00
.
0.00
.
3
0.45
0.75
0.66
0.85
0.00
.
0.00
.
4
0.50
0.75
0.63
0.84
0.00
.
0.00
.
5
0.60
0.69
0.00
.
0.00
.
3.94
0.63
6
0.33
0.83
0.32
0.99
0.00
.
0.00
.
7
0.24
0.87
0.36
0.99
0.00
.
0.00
.
8
0.31
0.98
0.27
1.07
0.00
.
0.00
.
9
0.30
0.90
0.34
0.99
0.00
.
0.00
.
10
0.36
0.87
0.50
0.99
0.00
.
0.00
.
11
0.41
0.90
0.41
0.99
0.00
.
0.00
.
12
0.50
0.79
0.45
0.85
0.00
.
3.95
0.02
13
0.38
0.75
0.00
.
0.00
.
1.14
0.78
14
0.25
0.99
0.00
.
0.00
.
0.00
.
15
0.26
0.99
0.00
.
0.00
.
0.00
.
16
0.34
0.87
0.65
0.98
0.00
.
2.05
0.35
17
0.25
0.99
0.00
.
0.00
.
0.00
.
18
0.27
0.98
0.00
.
0.00
.
0.00
.
19
0.70
0.81
1.85
0.89
0.15
0.98
0.48
0.98
20
0.21
0.99
0.00
.
0.00
.
0.00
.
21
0.27
0.98
0.00
.
0.00
.
0.00
.
22
0.57
0.69
0.00
.
0.00
.
4.79
0.98
23
0.00
.
0.64
0.86
3.77
0.69
0.00
.
24
0.65
0.86
0.98
0.69
0.00
.
0.00
.
25
0.55
0.74
0.80
0.86
0.00
.
0.00
.
26
0.21
0.98
0.28
0.99
0.00
.
0.00
.
27
0.43
0.99
0.82
0.99
1.55
0.89
0.79
0.88
28
0.26
0.98
0.29
0.99
0.00
.
0.00
.
29
0.24
0.98
0.67
0.99
0.00
.
0.00
.
30
0.48
0.74
1.15
0.69
0.00
.
0.00
.
31
0.45
0.73
0.95
0.75
0.00
.
0.00
.
32
0.55
0.69
0.93
0.69
0.96
0.79
0.55
0.79
33
0.52
0.87
0.88
1.01
0.00
.
0.00
.
34
0.22
1.10
0.36
1.12
0.00
.
0.00
.
35
0.25
1.09
0.00
.
0.00
.
0.00
.
36
0.37
0.89
0.00
.
0.00
.
2.98
0.74
37
0.00
.
1.92
0.74
0.00
.
18.61
0.40
38
0.29
1.09
0.00
.
0.00
.
0.00
.
39
0.26
1.09
0.00
.
0.00
.
0.00
.
40
0.29
1.09
0.00
.
0.00
.
0.00
.
147
CHAPTER 6 STANDARDIZED SOURCES
747
634
58¢
MARKETING DATA
21,715
16,773
599
OF
38¢
737
58¢
19,340
773
601
38¢
58¢
Kerry Gold brand
Regular 1 lb. 58¢
19,468
38¢
746
58¢
772
744
935
38¢
723
797
706
586
58¢
Brand B
Regular 1 lb. 49–54–58¢
6,460
5,927
2,960
1,256
1,713
971
1,359 1,268 1,239 1,513 1,127
49¢
54¢
13,908
49¢
1,355
58¢
1,689
44¢
49¢
1,235
58¢
1,184
49¢
1,021
58¢
1,450
58¢
5,441
2,462
54¢
44¢
1
8
April '85
790
54¢
15
857
596
578
529
6
May
13
20
44¢
22
29
1,159 1,224 1,260 1,058
50¢
1,559
474
50¢
Brand C
Regular 1 lb. 49–58¢
Brand D
Regular 1 lb. 54–58¢
6,174
611
58¢
2,653
1,742 1,810
49¢
4,886
50¢
8,650
981
38¢
58¢
10,014
2,319
1,318 1,076
1,246 1,231 976 1,017 1,188
753
675
683
10
17
1,810
532
54¢
49¢
27
3
June
Weeks
1,014
24
1
July
8
58¢
49¢
58¢
15
22
29
FIGURE 6-4
Unit sales per week (and prevailing price for the brand during the week).
TABLE 6-6
Share of Sales in
Nonpromotional
Weeks (%)
Share of Total Sales (%)
1st 12 yr
Brand
148
2nd 12 yr
(34)*
13
1st 21 yr
Kerry Gold
39
(7)
B
17
(8)
42
C
29
(13)
28
D
15
(11)
17
(12)
17.6
(66)
100
(72)
2nd 12 yr
18.9
19.3
(35)
30.4
30.3
(18)
33.1
33.9
100
16.5
100
*Figures in parentheses are deal merchandise shares of market. In the first half of the year, 66 percent of total unit
sales in the category were on deal.
thought they’d say no. Damn. Eighteen months down
the drain.”
“I know how you feel, but you have to understand
where the committee was coming from. It was a real
close call—as close as I can remember since I’ve had
this job. But the more carefully they considered your
tests results, the more it looked like the returns just
weren’t there.”
“Not there? All they had to look at was Appendix B
in my report—the data from Midland and Pittsfield: Sweet
Dream got a 3 percent share after 26 weeks. A trial rate of
15 percent. A repurchase rate of 45 percent. If national
performance were anywhere close to that, we’d have our
launch costs back in 14 months. Who can argue with that?”
“I’m on your side here, but I only had one vote,”
Bob said defensively. “We both knew what Barbara’s
position was going to be—and you know how much
weight she carries around here these days.” Barbara
Mayer was the Paradise group manager responsible for
established dessert products. She became a “grouper”
in 1985, after two enormously successful years as
LaTreat’s first product manager.
“And to be honest, it was tough to take issue with
her,” Bob continued. “What’s the point of introducing
Sweet Dream if you end up stealing share from
LaTreat? In fact, Barbara used some of your data
against us. She kept waving around Appendix C, griping that 75% of the people who tried Sweet Dream had
bought LaTreat in the previous four weeks. And repurchase rates were highest among LaTreat heavy users.
You know how the fourteenth floor feels about LaTreat.
Barbara claims that adjusting for lost LaTreat sales
means Sweet Dream doesn’t recover its up-front costs
for three years.”
LaTreat
Launched in 1983, LaTreat was the first “super premium” frozen dessert to enter national distribution. It
consisted of 3.5 ounces of vanilla ice cream dipped in
penuche fudge and covered with almonds. An individual bar sold for just under $2 and a package of four
was $7. Unlike LaTreat, which came on a stick, Sweet
Dream resembled an ice cream sandwich. It consisted
of sweet-cream ice cream between two oversized
chocolate chip cookies and coated with dark Belgian
chocolate. Its price was comparable to LaTreat’s.
Under Barbara Mayer, annual sales of LaTreat soon
reached $40 million, and it began making a significant
contribution to dessert groups profits. It accounted for
almost 5 percent of the market despite a price about
50 percent higher than standard frozen specialties.
Lately, however, competition had stiffened. LaTreat
faced tough challenges from three direct competitors as
well as several parallel concepts (like Sweet Dream) at
various stages of test marketing. The total frozen specialties market had grown fast enough to absorb these
new entrants without reducing LaTreat sales, but revenues had been essentially flat through 1986 and 1987.
Bill understood the importance of LaTreat, but he
was not the type to mince words. “You and I both
know things are more complicated than Barbara would
have people believe,” he told Bob. “There wasn’t the
same cannibalization effect in Marion and Corvallis.
And we never did a test in Midland and Pittsfield
where Barbara’s people were free to defend LaTreat. We
might be able to have it both ways . . .”
Bob interrupted. “Bill, we could stay here all night
on this. But what’s the point? The committee’s made its
decision. You don’t like it, I don’t like it. But these aren’t
stupid people. It’s hard to argue with the dessert group’s
batting average over the last five years. This may ring
hollow right now, but you can’t take this personally.”
“That’s easy for you to say,” Bill sighed.
“You know how this company works,” Bob
reminded him. “We don’t hold withdrawal of a new
product against the manager if withdrawal is the right
decision. Hell, it happened to me ten years ago with
that dumb strawberry topping. It made sense to kill
that product. And I was better off at the company for
it. The fact is, the committee was impressed as hell
with the research you did—although to be honest, you
may have overwhelmed them. A 40-page report with
30 pages of appendixes. I had trouble wading through
it all. But that doesn’t matter. You did a good job, and
the people who count know that.”
“I appreciate the sentiment, but that’s not why I
think this is the wrong decision. Sweet Dream is a go
on the merits.”
“Go home, play some golf this weekend,” Bob
counseled. “Things won’t seem so bleak on Monday.”
Bill never made it to the country club. Instead, he
spent the weekend worrying about his future at Paradise and puzzling over how the marketing committee
could have reached its no-launch decision.
Sweet Dream Proposal
Paradise Foods was a large, successful manufacturer of
packaged foods and household products whose markets were becoming increasingly competitive. Bill
believed that Paradise was vulnerable in this treacherous environment because of its failure to keep pace
with technological change—in particular, the increasing
sophistication of marketing research based on computer modeling, supermarket scanner data, and targetable cable television. Paradise certainly used these
tools, but to Bill’s way of thinking, top management
didn’t embrace them with the same enthusiasm as
other companies.
When Bill became produce manager for Sweet
Dream, he promised himself he would do a state-ofthe-art research job. The plan was to compare the
performance of Sweet Dream in two test markets
exposed to different advertising and promotion strategies. The campaign in Midland, Texas, and Pittsfield,
Massachusetts, struck an overtly self-indulgent tone—
“Go Ahead, You Deserve It”—and used limited price
149
promotion to induce trial. The campaign in Marion,
Indiana, and Corvallis, Oregon, emphasized superior
quality—“Taste the Goodness”—and used promotion
aggressively. Sunday newspapers in the two cities
frequently carried 50-cents-off coupons, and Sweet
Dream boxes included a 75-cent rebate voucher.
Bill used two computer-based research services—
InfoScan and BehaviorScan—to evaluate Sweet Dream’s
performance and long-term potential.2 InfoScan tracks
product purchases on a national and local basis for
the packaged-goods industry. It collects point-of-sale
information on all bar-coded products sold in a representative sample of supermarkets and drugstores. It
generates weekly data on volume, price, market share,
the relationship between sales and promotional offers,
and merchandising conditions. Bill subscribed to
InfoScan to monitor competitive trends in the frozen
specialties segment.
BehaviorScan is used in marketing tests to measure
the effect of marketing strategies on product purchases.
In a typical BehaviorScan test, one group of consumer
panelists is exposed to certain variables (that is, print or
television advertisements, coupons, free samples, instore displays), while other participating consumers
serve as a control group. Company analysts use supermarket scanner data on both groups of consumers (who
present identification cards to store checkout clerks) to
evaluate purchasing responses to marketing campaigns.
A typical BehaviorScan test lasts about one year.
Bill Horton’s research program had generated a
stack of computer printouts several feet high. He had
spent much of the spring trying to unravel the complex
interactions between different advertising and promotion strategies for Sweet Dream, the various promotion
deals Paradise was running on LaTreat, and the proliferation of other frozen specialties. Despite Bob’s advice
to relax, Bill spent Sunday afternoon in front of his
home computer, massaging the data one last time.
Competitive Analysis
On Monday, Bill arrived at his office a few minutes late.
He was surprised to find Barbara Mayer waiting on him.
“Sorry to drop in on you first thing,” she said, “but
I wanted to let you know what a fantastic job you did on
the Sweet Dream test. I’m sure you were disappointed
with the committee’s decision, and in a way I was too. It
would have been great to work together on the rollout.
But the data were pretty clear. We didn’t have a choice.”
“Well, I thought the data were clear too—but in the
opposite direction.”
“Come on, Bill, you can understand the logic of
the decision. The Midland and Pittsfield numbers were
fine, but they were coming at the expense of LaTreat.
There wasn’t so much cannibalization in Marion and
2
InfoScan and BehaviorScan are actual services offered by Information Resources, Inc.
150
Corvallis, but the Sweet Dream numbers weren’t as good
either. Trial was acceptable, but repurchase was low. We
might make money, but we’d never meet the hurdle rate.
Every so often a product just falls between two stools.”
“So we’ll do more tests,” Bill countered. “We can
play with the positioning in Marion and Corvallis. Or
we can start from scratch somewhere else. I can have
us wired to go in three weeks.”
“We’ve already taken 18 months on Sweet Dream,”
Barbara said. “The committee felt it was time to try
new concepts. I don’t think that’s so unreasonable.”
“You’re forgetting two things,” Bill replied. “First,
with freezer space as tight as it is, the longer it takes to
come up with another product, the harder the stores are
going to squeeze us. Second, other people are going to
find out how well Sweet Dream did in Midland and
Pittsfield. We’re the only ones who get the BehaviorScan numbers, but you know the competition is
monitoring our tests. What do you think Weston &
Williams is going to do when it sees the results? It’ll
have a Sweet Dream clone out in a few months if we
don’t launch.”
Weston & Williams (W & W) was a leading supplier of household products that was diversifying into
foods, including desserts. It had a reputation as a
conservative company that insisted on exhaustive
prelaunch research. But the trade press recently had
reported on W & W’s decision to rush Pounce—a
combination detergent, colorfast bleach, and fabric
softener—to the market on the basis of very preliminary tests and data from a competitior’s test markets.
W & W had thus become the first national entrant in
the “maxiwash” category.
“Bob made that argument Friday,” Barbara said.
“But you can guess how far he got. The guys upstairs
have a tough enough time taking our own computer
data seriously. They don’t buy the idea that someone
else is going to jump into the market based on our tests.
Plus, that would be a huge risk. Pounce may have
given Weston & Williams all the gray hair they can
stand for a few years.”
“From what I can tell, Barbara, the only issue that
counted was cannibalization.” Bill’s voice betrayed a
rekindled sense of frustration. “I understand you want
to protect LaTreat. I understand the company wants to
protect LaTreat. But it seems to me we’re protecting a
product that is getting tired.”
“What are you talking about?” Barbara objected.
“Profits aren’t growing as fast as they used to, but
they’re not dropping either. LaTreat is solid.”
Sales Analysis
“Come on, Barbara. Your people have really been promoting it in the last two quarters—shifting money out
of print and TV and into coupons and rebates. Total
spending hasn’t changed, so profits are OK. But LaTreat
has gotten hooked on promotion. And all the wrong
Millions of dollars
$50
$40
$30
$20
$10
0
Nominal sales
Average sales
Sales without promotion
Apr May
1987
June
July
Aug
Sept
Oct
Nov
Dec
Jan
1988
Feb
Mar
FIGURE 6-5
LaTreat sales with and without promotions (seasonally adjusted figures presented
on an annualized basis).
kinds of promotion. You’ve got people accelerating
future purchases and price-sensitive types jumping in
whenever LaTreat goes on sale. Who needs that?”
“Where are you getting this stuff?” Barbara
demanded, “I didn’t see it in your report.”
“I spent the weekend running some more numbers.” Bill replied. “Take a look at this.”
Bill punched a few buttons on his computer keyboard and called up a series of graphs. The first documented the growing percentage of LaTreat sales
connected with promotional offers. A second graph disaggregated LaTreat’s promotion-related sales by four
buyer categories Bill had created from BehaviorScan
data. “Loyalists” were longtime customers who
increased their purchases in response to a deal. “Trial
users” bought LaTreat for the first time because of the
promotion and seemed to be turning into loyal customers.
“Accelerators” were longtime customers who used
coupons or rebates to stock up on product they would
have bought anyway. “Switch-on-deal” customers were
nonusers who bought LaTreat when there were promotions but demonstrated little long-term loyalty. Bill’s
graph documented that a majority of LaTreat’s coupon
redeemers fell into the last two categories, with “loyalists”
accounting for a shrinking percentage of sales.
Finally, Bill called up his ultimate evidence—a
graph that adjusted LaTreat sales to eliminate the effect
of promotions. (See Figure 6-5.)
“I’m amazed you spent your weekend doing this,”
Barbara said, “but I’m glad you did. It’ll help us think
through future marketing strategies for LaTreat. But it
doesn’t change what the committee decided. It’s time
to move on.”
“I’m not so sure,” Bill replied. “I hope you don’t
mind, but I think I should show these data to Bob.
Maybe he can convince the committee to reconsider.
After all, if LaTreat is weakening, its going to show up
in your profit figures sooner or later.”
“Data don’t make decisions, Bill, people do. And the
people on the marketing committee have been in the
industry a lot longer than you. Their gut tells them
things your computer can’t. Besides, you and I both
know that when you collect this much data, you can
make it show just about anything. Go ahead and talk to
Bob, but I’m sure he’ll see things the same way I do.”
Assignment
1.
2.
What (if any) issues did Bill Horton overlook in
evaluating Sweet Dream?
Do you think Sweet Dream should be launched?
Give the rationale for your answer. Is there any
additional research you feel should be conducted
before a final decision on the Sweet Dream launch
is made?
151
7
MARKETING RESEARCH ON THE
INTERNET1
Learning Objectives
■ Give a brief overview of the past and current developments of the Internet.
■ Describe the components of the Internet.
■ Illustrate current commercial research applications.
■ Demonstrate the current use of the Internet as a marketing research tool.
■ Provide an outlook of future developments.
The Internet is arguably an immensely important tool that has added a new dimension to traditional marketing research intelligence. With volumes of relevant information available at the click
of a mouse, decision making has assumed a new perspective. The number of Internet users around
the world is constantly growing. According to data published by comScore Networks, the Internet
currently reaches 747 million people worldwide. The highest growth percentages over the past year
comes from developing economies—namely India (33 percent), the Russian Federation (21 percent)
and China (20 percent).2
At the start of the new century, 55 million of us were shopping to our hearts’ content at
home. Shopping online, and researching products online, has increased since 2000. In 2000, while
22 percent of Americans had bought a product online, this rate grew to 49 percent in 2007. For
product-related research, 35 percent had used the Internet in 2000, and 60 percent in 2007.3
According to a survey conducted by Forrester Research Inc., total retail online spending in general merchandise will reach $151.3 billion by 2010, up from $75.7 billion in 2004. Those numbers
exclude online ticket sales projected by Forrester at $9.4 billion in 2010, up from $4.3 billion in
2004. Further, the report projects additional online sales of about $155.5 billion in 2010 from the
combined category of automobiles and auto parts, food and beverages and travel, up from $65.6
billion for the category in 2004.
Shopping on the Internet during the holiday season has come a long way since late 90s. Sales at ecommerce sites for retail spending in 2006 reached $102.1 billion, a 24 percent increase over 2005 sales.
The holiday shopping season (November 1 through December 31) brought in $24.6 billion, up 26 percent from 2005. The e-commerce spending summary for the period between 1998 and 2006 is given in
Table 7-1.
CURRENT TRENDS IN WEB USAGE
152
National trends of Internet usage have increased with improved education, income, and the presence of school-age children at home. According to the U.S. Census Bureau 2009 population survey,
69 percent of American households had access to the Internet at home in 2009, more than quadruple the percentage in 1997. In a report published by Pew Internet and American Life Project in
2005, 68 percent of adults were found to use the Internet, up from 63 percent last year. It
further reported that 22 percent of American adults have never used the Internet. Advertisers are
taking advantage of increased Internet use, with Internet advertising revenue at $26 billion in 2010,
more than triple the amount from five years ago. The Census report found that in 2008, 78 percent
CURRENT TRENDS
IN
WEB USAGE
TABLE 7-1
E-Commerce Spending Summary, 2000–2011
Retail Sales (millions of
dollars)
Total
E-Commerce
E-Commerce
as a Percent
of Total
2011
3,123,715
141,727
4.5%
2010
3,887,725
166,513
4.3%
2009
3,639,467
144,462
4.0%
2008
3,945,364
142,508
3.6%
2007
4,008,096
137,564
3.4%
2006
3,885,877
114,341
2.9%
2005
3,693,913
92,505
2.5%
2004
3,467,850
73,920
2.1%
2003
3,266,790
57,931
1.8%
2002
3,135,269
45,007
1.4%
2001
3,067,965
34,448
1.1%
2000
2,984,979
27,574
0.9%
SOURCE: Compiled from Annual Retail Trade Surveys, U.S. Census Bureau
of adult Internet users purchased products or services online, up from 2 percent in 1997.4 As more
customers rely on the Internet for their daily transactions, analysis of this becomes essential for
businesses in order to determine the profile of customers, customer lifetime value and effectiveness of promotional campaigns, among others. The Internet usage today is witnessing a change
among the older users. eMarketer estimates that in 2005 there are 33.2 million people online in the
United States between the ages of 50 and 64, triple the number of 65 online users.5 While a majority of the e-commerce market is yet to design senior-specific sites, some pharmaceutical and financial sites have developed websites usable by everyone and particularly catering to older people.
Pfizer Inc. has websites that are specifically designed to aid older people by providing information
on senior citizen products. Financial companies like Fidelity Investments and Charles Schwab Corp.
have also incorporated usability features for seniors on its websites.6
The prominence of the Internet is being increasingly witnessed in developing countries as well.
According to Internet World Stats, North America is home to about 13 percent of world Internet
users, with a penetration rate of 78.3 percent. However, it recorded the lowest usage growth of
151.7 percent between 2000 and 2010. During the same period, the highest usage growth of 2,527
percent was registered in Africa. The Middle East was second, recording over 1987 percent growth.
The Latin America/Caribbean region was in third place with 1037 percent growth.7
Language too plays an important role in Internet usage. According to a report published by
Global Reach in 2002, Internet users by native speakers were found to be English—26.8 percent, Chinese—24.2 percent, Japanese—9.7 percent, Spanish—7.2 percent, German—6.7 percent, Korean—4.5
percent, Italian—3.8 percent, French—3.5 percent, Portuguese—3.0 percent, Russian—2.9 percent,
Dutch—2.0 percent.8 With such a wide variety of users taking to the Internet, marketers are incorporating language selections in their home pages to cater to all speakers.
It is quite evident that as the Web usage continues to grow, it provides marketers with a
whole variety of options in terms of reaching a wider audience and better exposure to the products or services that are offered. In this context, monitoring Web usage assists Web developers
in developing a website that is accessible to a larger population.
The Internet offers real advantages to all firms, including small ones. Since the cost of constructing a website is reasonably low, a firm with less than a million dollars in sales is as accessible
to potential customers worldwide as Nestle or French Telecom. The Internet can “level the playing
field” for marketing to potential customers through efficient use of the medium. Examples abound
of companies that have made it big through the use of the Internet. According to research conducted
153
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
by Forrester Research, e-commerce transactions in 2006, 2007, 2008, 2009, and 2010 are expected to
reach $198 billion, $228 billion, $258 billion, $288 billion, and $316 billion, respectively. The estimates
include travel, typically the biggest e-commerce sales driver. It further estimates that by 2010,
e-commerce will have a 14 percent compounded annual growth rate over six years. Excluding travel,
e-commerce was expected to grow 14 percent in 2005 from 2004.9
WWW INFORMATION FOR MARKETING
DECISIONS
Online Advertisements
With technological advancements and the rapid spread of the World Wide Web, advertising
online provides instant access to the fastest growing media audience in the world. Online advertising provides an interactive element like no other traditional advertising medium. The origins
of online advertising can be found in the banner advertisements first featured in the early 1990s.
The forms of advertising soon expanded to opt-in e-mails, search engine advertising, floating
animated pages, and audio and video streams among others. Since online advertising provides
an interactive element, it enables advertisers to develop a one-to-one relationship with prospective and existing clients. With the advertisements carrying links to the advertiser’s website,
measurement and tracking of advertisements become easier. The top five companies ranked by
Sponsored Link Impressions in August 2008 are listed in Table 7.2.
One of the popular methods of online advertising is Search Engine Advertising. According to a
research conducted in January 2005 by Pew Internet & American Life Project, nearly 87 percent of
the viewers have had successful search experiences. It further says that 84 percent of Americans who
are online have used search engines, and, on any given day, 56 percent of people online are using
them.10 With search sites becoming popular, search engine advertising can help advertisers reach
their target audience in more ways than one. This type of advertising involves advertising on the
Web by tracking keyword searches. In other words, when a viewer is searching for information on
the Internet, the keywords used for the searches would lead the viewer to the advertiser’s website,
amid other search results. For this to be effective, the target audience must be properly identified
and the keyword phrases used for searching must be placed strategically on the website.
Advertising in local media sources can be an effective way to overcome the targeting issue.
According to the research report “What Local Web Sites Earn: 2005 Survey” by Borrell Associates,
TABLE 7-2
Top Companies by Sponsored Link Impression
as of May 2009
Company
Impressions (in millions)
InterActiveCorp
768.4
Research In Motion Ltd.
662.2
eBay Inc.
625.6
Yahoo Inc.
451.3
Amazon.com Inc.
363.9
Bank of America Corp.
319.8
Infospace Inc.
281.7
e-Unity Corp.
249.0
Target Corp.
238.4
SOURCE: http://searchenginewatch.com/article/2048728/Advertising-Placements-by-Industry-and-Top-Sponsored-Links-March-2009
154
WWW INFORMATION
FOR
MARKETING DECISIONS
TABLE 7-3
Projected U.S. Local Online Ad Spending, 2003–2005
Spending Rank
2004
2005
Local Online Ad Spending (in $ millions)
District Metropolitan Area
2003
2004
2005
1
1
New York City
$137.15
$184.09
$240.49
2
2
Los Angeles
$ 85.90
$106.67
$141.32
3
3
Chicago
$ 59.29
$ 76.38
$100.58
7
4
Washington, D.C.
$ 45.69
$ 54.09
$ 91.83
5
5
Philadelphia
$ 52.78
$ 69.17
$ 91.03
SOURCE: Borrell Associates Inc., 2005 and Ad Audit Services, 2005.
Adapted with permission from Sean Michael Kerner, “Online Ad Spend Swells,” clickz.com, April 1, 2005,
http://www.clickz.com/clickz/stats/1707269/online-ad-spend-swells
local online spending is expected to hit $3.9 billion in 2005, an increase of 46 percent from 2004.11
Table 7-3 lists the top five metro areas for local online spending. The survey listed newspapercontrolled websites as the leading category for online advertisement revenue in 2004, accounting for
$1.19 billion; about 44 percent of all local online advertising spending. Internet pure-plays such as
AutoTrader.com, Monster.com, and HouseValues.com accounted for 40 percent of online advertising spending. The remaining 16 percent was split between yellow pages/interactive directories
(6 percent), paid search companies (5 percent), TV stations (4 percent), and radio stations (1 percent).
The report attributed the low share of radio stations as compared to television stations in online
advertisements to increasing use of broadband and MP3 players. The growth of these players has
given rise to nontraditional ways of audio broadcasting. However, the reporters appeared hopeful
that television and radio stations will catch up through new revenue streams.
A relatively new online advertising option is contextual advertising. This form of advertising
goes one step further from search engine advertising. A contextual advertisement is similar to the
advertisement that is displayed at the top of search results, but is instead displayed on the content
page that contains the relevant material.12 For instance, social-networking Web site Facebook Inc.
is working on a new advertising system that would let marketers target users with ads based on
the information people reveal on the site about themselves. Eventually, it hopes to refine the system to allow it to predict what products and services users might be interested in even before they
have specifically mentioned an area.13
In certain cases, participating content sites achieve high search engine rankings. Limiting the
advertisements to these sites may not be guarantee maximum viewership. This is because,
searchers may not click these search result sites. Hence, advertising in sites that fit the context of
keyword may not only bring in more viewers but will also capture those who do not visit through
a search engine.
Online Tracking Measurements
An average consumer through exposure to online advertisements consumes enough to call for study
and monitoring. A true measurement of an advertisement’s performance is thus essential for creating and retaining customers. Before the advent of the Internet, researchers relied primarily on telephone surveys to ascertain the advertisement recall rates by interviewing a sample. However, with
the penetration of the Internet, online tracking of advertisements has become a viable option.
Online tracking is advantageous to researchers in many ways. First, Internet tracking offers
researchers a means to test viewers’ advertisement recognition by providing visual cues, unlike
telephone surveys that only provide verbal cues. Second, online tracking yields faster results.
This is because everyone in the sample is sent an invitation to fill out the response online. This
saves a lot of time and energy as compared to the telephone version. Finally, not only is it now
155
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
MARKETING RESEARCH IN ACTION 7-1
Tracking Flu—The Google Way
Ever wondered what happens to all the search words in a search box? Google, a leader in web search, has
been tracking such key words and offering search related content apart from the search results. Google has
extended this concept and has launched the Google Flu Trends. This new tool (www.google.org/flutrends)
from Google allows web users to see the flu activity across the United States. By tracking the search phrases
related to “flu symptoms,” Google has been able to detect the regional outbreaks of the flu a week to 10
days before they are reported by the U.S. Centers for Disease Control and Prevention. This tool analyzes
the searches as they happen, and creates graphs (both at a state and country level) that, ideally, will show
where the flu is spreading.
Talking about the tool’s predictive accuracy, Google says it witnessed a spike in the search data related
to flu symptoms about two weeks prior to the flu report released by the CDC. Since the CDC relies on
the health care providers to provide information related to diseases and their outbreaks, valuable time
is lost in data collection efforts. Google on the other hand, collects data as it happens and tracks the flu
spread. This tool could help accelerate the response of doctors, hospitals, and public health officials;
reduce the spread of the disease; and, potentially, save lives.
SOURCE: Miguel Helft, “Google uses web searches to track flu’s spread,” New York Times, November 11, 2008.
156
possible for researchers to know the viewer awareness of advertisement campaigns, they can also
determine the degree of awareness of a campaign by comparing segments of the audience. With
traditional telephone recall studies, only determining advertisement awareness was possible.14
In order to track online advertisements, new methods are being incorporated by large companies. For instance, Yahoo! and A.C. Nielsen are using a research methodology called Yahoo!
Consumer Direct for measuring the impact of online advertisements on off-line retail purchases of
consumer products. With the Internet undergoing change frequently, the profile of viewers changes
accordingly. While new metrics for tracking online advertisements are being implemented, marketers do not have a “perfect metric” for tracking.15 Thus, marketers will have to recognize the
changing consumer profile and develop suitable tracking methods. Marketing Research in Action
7-1 presents a novel use of online tracking practices.
Some examples of firms involved in keeping track of online trends which can provide valuable information to companies for marketing decision making include the following.
Cyberatlas, a division of I/PRO, compiles and publishes public-domain information about
Internet market size, growth, forecast, and demographics. In addition, Cyberatlas provides news
on WWW markets such as advertising, finance, and search services, as well as normative data
and industry benchmarking studies. The domain name market shows a strong rate of new registrations and a stabilizing base of domain names. According to “The Domain Name Industry Brief,
March 2007” published by VeriSign registered domain names have grown to a base of 120 million worldwide across all top-level domains. The total number of domains comprises a 32 percent
increase between 2005 and 2006.16
Forrester Research from Cambridge, Massachusetts, offers a service called New Media Research.
One component of this project is Media & Technology Strategies, which focuses on new business
media models for publishers, broadcasters, and information service providers. Research topics
include, for example, how information will be priced in the electronic marketplace. This is critical information for corporations which do heavy advertising on the Web (such as IBM or
Netscape), because they want an adequate return on their advertising spending. Many different
proposals have been made for pricing ads on the Internet. Forrester found out that advertisers
complain about pricing schemes based on cost per thousand and would rather see click-through
(based on the number of clicks for the ads) pricing. Table 7-4 shows the top Web advertisers of
2009 and 2010.
Global online advertising expenditures in 2004 were $769.2 million; up from $654.6 in 2002.17
The advantages of advertising on the Internet cannot be understated. First, a targeted audience can
be reached. Many websites request personal information about the visitor, which can be used not
WWW INFORMATION
FOR
MARKETING DECISIONS
TABLE 7-4
Internet Ad Revenues by Major Industry Category, 2009 vs 2010
20%
Retail
21%
2009
2010
16%
Telecom
13%
12%
Financial Services
12%
11%
Auto
11%
10%
Computing Products
8%
6%
Consumer Packaged Goods
7%
6%
Leisure Travel
7%
Pharma & Healthcare
4%
5%
Media
4%
4%
Entertainment
4%
4%
SOURCE: http://www.onlinemarketing-trends.com/2011/08/worldwide-internet-advertising-to-reach.html
*% of total
only to define the site’s audience but also to produce customized ads to be shown on the next visit.
Many sites are, in fact, a mechanism for delivering an audience rather than selling a product or a
service. Second, Internet advertising is very much like direct-response advertising in that customers
can link directly to an advertiser’s home page and request information, browse a catalog, buy a
product, or order a service.
One problem in Web advertising is audience measurement. A firm would like to measure
“eyeballs,” that is, how many total exposures an ad receives. However, on a given site visit, customers often go back and forth rapidly between hypertext pages. Measuring the number of times
a particular page is on the screen may not be the appropriate measure of exposure, as the screen
time may have been very brief. A minimum amount of time might be necessary before it can
count as a “hit.” Or the viewer may not scroll all the way down the screen. These issues confront
the large number of companies interested in measuring site audiences.18 To address these problems, some new services are being developed.
AdCount by NetCount is a tracking service of Web users that attempts to measure the effect
of Web ads. It promises not only to monitor how many Web cruisers see an ad but
whether they click on it to be connected to the advertiser’s own website. Such “clickthrough” rates are the true measure of whether an ad placement had an effect. For this
157
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
purpose, special software is installed on the host computer. The software sends frequent
updates of ad activity on the sites to a central computer from which the reports are generated. These reports usually comprise daily and weekly totals of ad impressions and
click-throughs, daily and weekly totals of ad transfers (clicks on ads which successfully
deliver the user to the advertiser’s website), and advertiser summary reports with comparable statistics on each ad tracked.
Some good examples of how companies are using the Internet and the World Wide Web are the
following (source: World Wide Web):
The Ford Motor Company is now offering an online store that allows users to buy cars of their
choice online. Users can select the brand, model, and the features for the car. In short, they
can build their car online and get the estimated cost. They could apply for a financial loan
as well. (www.ford.com)
Amazon.com opened its virtual doors in July 1995 with a mission to use the Internet to transform book buying into the fastest, easiest, and most enjoyable shopping experience possible. Today, Amazon.com is the place to find and discover anything you want to buy online.
It is the leading online shopping site, with 12 million people in more than 160 countries
shopping with them. They have Earth’s Biggest SelectionTM of products, including free
electronic greeting cards, online auctions, and millions of books, CDs, videos, DVDs, toys
and games, and electronics. (www.amazon.com)
MapQuest.com is a leader in Advanced Mapping Solutions. Since 1996, MapQuest.com has been
offering the number one website for consumers to get online, interactive maps and driving directions. MapQuest.com is also a leading map and driving directions “enabler” of
business websites, corporate intranets, and other software applications. MapQuest. com
offers over 30 years of cartographic, digital mapping, and mapping software applications,
and, most recently, advertising and Internet experience. During January 1999, MapQuest
delivered over 70.3 million maps and over 13.5 million driving directions through its own
website and through third-party websites. (www.mapquest.com)
eBags is the world’s largest online retailer of bags and accessories for business, sports, and
travel. They carry a complete line of premium and popular brands, including Samsonite,
JanSport, North Face, Eagle Creek, and Timberland. From backpacks and carry-ons to
computer cases and adventure gear, eBags combines the best selection of products with
unrivaled services and extremely competitive prices. It is five times larger than most
department stores. eBags allows you to browse a huge selection of bags in a large variety of colors. Unlike most stores, eBags offers smaller specialty brands for those looking
for something unique. In addition, eBags carries bags in every price range, ensuring that
shoppers find the perfect bag at the perfect price. (www.eBags.com)
THE INTERNET AND MARKETING RESEARCH
TODAY
158
Over the past few years, online research has become an important mode of quantitative market
research. The newsletter Inside Research quotes that online market research spending has
increased to an estimated $1.3 billion in 2006 from $253 million in 2000. This meteoric rise in
online presence has been largely due to the wide acceptance of the Internet as a source of information. Other factors include the wide presence of Internet and the declining response rates of
the telephone/CATI method. While the online mode of research is clearly more cost effective
than the traditional modes of research, the quality of online research results is still a topic for
debate. The rapid advancement of the online mode has given rise to many market research innovations such as opt-in panels like Harris Interactive and Greenfield Online, self-service survey
THE INTERNET
AND
MARKETING RESEARCH TODAY
TABLE 7-5
Uses for Online Services
Activity
Buy or make a reservation for travel
Rural Internet Users (%)
Urban & Suburban
Internet Users (%)
51
65
Online banking
34
43
Online classifieds
30
37
Read a blog
21
28
Download screensavers
28
22
Download computer games
25
20
Class for credit
15
11
Fantasy sports
9
7
SOURCE: Pew Internet & American Life Project. Enid Burns, “Rural America Slow to Adopt Broadband,” clickz.com.
February 27, 2006, http://www.clickz.com/clickz/stats/1705120/rural-america-slow-adopt-broadband#table2
providers such as Zoomerang and SurveyMonkey.19 Like any traditional information resource,
the Internet has certain advantages and disadvantages. Besides this, the Internet is characterized
by very dynamic technological developments, which in turn influence the information search
process. This section looks at a variety of ways the Internet can be utilized for marketing research.
Primary Research
The Internet is now being used for collecting primary information for use in mainstream marketing, though with caution. The various ways of collecting information are briefly discussed.
E-mail Surveys. A familiar example for this is the use of electronic mail for survey research. This
technique uses e-mail for the entire process of receiving, completing, and returning questionnaires. This type of survey technique has a number of advantages:
■ Greater speed of delivering electronic mail over regular postal mail. Questionnaires are delivered, or redelivered if lost, in a matter of seconds.
■ Higher speed of delivering responses and feedback.
■ Cost-savings benefits over regular mail surveys.
■ No intermediaries—e-mail messages are usually read only by the recipient.
■ Asynchronous communication. Unlike telephone surveys, messages can be sent, read, and
replied to at the convenience of the user.
While e-mail survey is an important and effective tool, it is not without its share of problems.
Important among the drawbacks of e-mail surveys is spamming or sending unsolicited bulk messages. Spam in e-mail started to become a problem when the Internet was opened up to the general public in the mid-1990s. It grew exponentially over the following years, and today comprises
a majority of all the email in the world. A market research firm, Radicati Group Inc. reports that
the average number of e-mail messages sent daily in 2010 was around 294 billion, of which 89%
are spam. Of these junk messages, nearly 50% were delivered. While mail filtering programs are
being developed at a fast pace to curb the spam menace, spammers are finding newer ways to
send messages to e-mail users.
Some of the less desirable properties of e-mail surveys are that the security of electronic mail
is low in comparison to traditional media. Also, even though many researchers guarantee
anonymity of participants, respondents may often not answer truthfully for fear that their identity might be revealed somehow. As the population of the Internet and online users increases, new
159
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
MARKETING RESEARCH IN ACTION 7-2
GVU’s WWW User Survey Methodology
Unlike most other surveys, GVU’s WWW users’ surveys are conducted on the Web (that is, participants
respond to questionnaires posted on the Web). In fact, GVU pioneered the entire field of Web-based
surveys. The GVU Center conducts the surveys every six months as a public service to the WWW community.
The GVU surveys employ nonprobabilistic sampling. Participants are solicited in the following
manner:
■ Announcements on Internet-related newsgroups (for example, comp.infosystems, www.announce,
comp.internet.net-happenings, etc.)
■ Banners randomly rotated through high-exposure sites (for example, Yahoo!, CNN, Excite,
Webcrawler, etc.)
■ Banners rotated through advertising networks (for example, Double Click)
■ Announcements made to the www-surveying mailing list, a list maintained by GVU’s WWW User
Surveys composed of people interested in the surveys
■ Announcements made in the popular media (for example, newspapers, trade magazines, etc.)
There are several points to be made here. First, the above methodology has evolved due to the fact there is
no broadcast mechanism on the Web that would enable participants to be selected or notified at random.
As such, the methodology attempts to propagate the presence of the surveys through diverse media. Second, high-exposure sites capture a significant portion of all WWW user activity as measured by PC-Meter.
These sites are specifically targeted to increase the likelihood that the majority of WWW users will have
been given an equal opportunity to participate in the surveys. Additionally, content-neutral sites are chosen from the list of most popular sites to reduce the chance of imposing a systematic bias on the results.
Finally, the Seventh Survey was the first survey to experiment with the random rotation of banners through
advertising networks. The ability of the advertising networks to randomly rotate banners is relatively new,
one that did not really exist during the first three years of GVU’s Surveys. This ability goes a long way
toward ensuring that members of the WWW community have been selected at random. Since this technique
is still quite experimental, its effect on the reliability of the results cannot as yet be determined, though it
will continue to be examined in the future research.
SOURCE: “GVU’s 10th WWW Survey Results,” http://www.cc.gatech.edu/gvu/user_surveys/survey-1998-10
research issues have arisen concerning the demographics and psychographics of the online user
and the opportunities for a product or service. Marketing Research in Action 7-2 gives an example for how a Web survey can be conducted. The key issue is to provide a disclaimer that the
firm seeks only composite data, and that names and addresses will not be used for solicitation or
sold to another vendor. This kind of survey can also be conducted by announcing it on appropriate lists and groups, and then, in response to a request, sending the survey out via e-mail.20
Interactive Forms. Primary data collection may also be performed through interactive forms which
are filled out on the screen. For example, in order to gain specific demographic information and reading habits about Internet users, one company that provides access to newspapers and magazines ran
a contest for those willing to fill out an online survey. A news release was put out to the nethappenings mailing list, announcing that the information was being made available on their Gopher.
160
Online Panels. Online panels are an effective way of conducting Internet research whereby
groups of people agree to participate in surveys and exchange their views. Companies such as
Harris Interactive and Synovate conduct online panels with participants from all over the world.
The information collected from the members through these panels is used for developing
business and marketing solutions. Apart from being cost effective and fast, online panels provide high-quality data acquired from willing, interested, and motivated participants. As they
THE INTERNET
AND
MARKETING RESEARCH TODAY
provide anonymity and convenience to the panel members, it provides a viable option for themarketers to research sensitive topics. Some of the other advantages of online panels include the
possibility of covering a wide audience, the ability to conduct international research from
one place, and the viability of electronic monitoring of respondents through log file analyses.
The quality of online panels is determined by recruitment, maintenance, and response rate
standards.21 The online panel recruitment can be broadly classified into two categories: actively
recruited online panels and passively recruited online panels. In case of actively recruited online
panels, participants are carefully selected and targeted, either off-line or online. Since this
involves a selection procedure, it becomes a closed pool of participants. In passively recruited
online panels, participants find their way to register themselves through the various links available on the Internet and in magazines. Since this does not involve any selection procedure, it
functions as an open pool of participants. The recruitment of online panel members may also
include employees and customers. The point to be noted here is that, recruiting members for
the online panel based on the purpose of the online panel would yield better results from the
subsequent surveys.
The maintenance of online panels is essential in conducting online research. This calls for a
continuous monitoring of the panel members. Researchers, while monitoring the existing panel
members, must also be on the lookout for new panel members to replace users who have left
the panel. The existing panel members may also be replaced with new members because of their
familiarity with research methods, which may lead to response bias. The process of recruiting
new members is easier in online panels than in off-line panels, as researchers can stay in constant touch with the members and responses can be obtained instantly. Maintaining online panels
also calls for communicating with panel members on a regular basis, ensuring the time relevance
of collected data and satisfaction of panel members, as this would be a good indicator of panel
quality.
Response rates perform an important part in online research by indicating the quality of survey research and in determining the cost of the project. When respondents “opt in” to the panel,
they bring in a significant amount of information relating to demography and behavior. The information thus collected later helps in deciding the samples. Opt-ins typically lead to higher response
rates as the questionnaire can be shorter. This is possible because the researcher would already have
the basic information of the panel members and need not spend time and energy on collecting the
information again. The response rates also depend on the length of online survey, the amount and
type of incentives offered, and the technical equipment available to the respondents. Researchers
have to be aware that having e-mail addresses or having a large panel does not translate into a
higher response rate.
With business problems demanding attention, further limited by time, online panels offer
researchers and marketers a way to obtain immediate results from the targeted consumers and
at their convenience. Given the higher response rates, online panels are gaining popularity
among researchers. When compared to other traditional methods such as telephone and mail
studies, The Council for Marketing and Opinion Research reports that cooperation rates for Internet panel studies in 2010 were as high as 65 percent.
Online Focus Groups. Online focus groups are conducted entirely online—everything
from recruitment and screening (which the recruiter does via e-mail) to moderation of the discussion
itself. This method allows researchers to reach target segments more effectively. Online focus groups,
while lacking the dynamics of a face-to-face discussion, provide a unique alternative to the traditional method. An online environment allows respondents to interact voluntarily “behind their computer screens,” and therefore encourages them to respond with honest and spontaneous answers.
Furthermore, all respondents, extroverted and introverted, get a fair chance to express their views,
and “instant messaging” allows the key focus-group players to interact privately with each other.
Although online respondents do not “hear” each others’ answers, they can see them. Thanks to
“emoticons” (the use of certain keys that, typed in combination, look like facial expressions), cyberspace allows online focus group participants and moderators to express themselves.22
161
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
As the online population increases, the demographics broaden, enabling remote global segments to be reached, something not possible via traditional methods. One of the limitations to
online research is that the results cannot be projected to the general population because not everyone has access to a computer, modem, and online service. Another difference between online and
traditional qualitative research is that cyberspace is populated by trend leaders. Commonly targeted by marketers, advertisers, and product manufacturers, trend leaders are early adopters
who try out new ideas, products, services, and technologies before these innovations reach popularity in the mass market.
Companies are increasingly collecting information from their website visitors. Especially for
companies which sell over the Web, collecting information about potential customers who have
Internet access is critical. This type of data collection can serve a number of purposes:
■ Counting and describing website visitors in order to customize website content to suit their needs
■ Collecting additional information for customer databases, which then may be used by product development, sales, marketing, or service departments
■ Receiving questions or suggestions regarding the use of a product
■ Receiving and answering complaints
It is surprising to see how many website visitors are willing to leave personal information (such
as address, phone number, e-mail number). In many cases, companies offer a small reward as a token
for the complying visitor. For example, upon completing a questionnaire at the Campbell’s Soup website, the company mails a number of coupons for their products to the customer. Other tokens from
packaged-goods companies are good for product samples, price drawings, or sweepstakes.
I/PRO has launched a service called I/CODE that lets users submit information about themselves once; it then makes the data available to sites participating in the I/CODE scheme. The advantage to the user is being able to skip answering the same questions repeatedly, and I/CODE offers
prizes as an incentive to register. Publishers and advertisers pay for the information, and in return
can build a database of users and track their behavior. C|NET gathers information about users without asking questions. Its servers can recognize each visitor’s browser and computer type as well as
their domain (.com, .edu, .org, etc.). It uses this information to deliver a “banner” (advertising) tailored to different types of users—for example, a Mac warehouse banner for a Mac user, and a PC
warehouse ad for a PC user.
Return on Investment (ROI) is becoming increasingly popular as a new technology in helping promote direct marketing. ROI tracks the consumers that are surfing online who click on the banner ads.
The positive trend for ROI stems from the demand for quantifiable results from business investments
made by executives and shareholders. This has resulted in an increasing number of firms to equip
themselves with analytic software programs that measure ROI on a real-time basis. A recent report on
marketing ROI and ROI measurement trends found that in 2005, 78 percent marketers felt their budgets for marketing analytics was underfunded. In 2006 that percentage had dropped to 64 percent.23
The click-through rates measure the type of consumer response by the amount of clicks produced. In
counting returns, ROI becomes a key metric in evaluating the effectiveness of banner ads for some
online advertising agencies. The calculation on the return of ad spending to track actual sales, software downloads, or user registration is done in the following way:
■ A banner ad is served to an online viewer, who is tagged with an identification code stored
in the browser’s cookie.
■ The user will then click on the ad and is sent to the link promoted by the advertiser.
■ An ad-tracking firm serves a tiny pixel-sized image, which is posted on key pages of the website the advertiser wants to track. Such pages often begin software downloads or are the
“thank you” pages that are shown after a consumer makes a purchase.
162
■ When the user clicks on the key page, identification information stored on the cookie is sent
to a central server that tracks the pixel-sized image and the resulting consumer action.
THE INTERNET
AND
MARKETING RESEARCH TODAY
■ Results logged at the central server are compared with overall investment in advertising to
generate return on investment results.24
User Generated Reviews. Listening to customer reviews is becoming an increasing source of vital
marketing information. User generated information, especially product and site reviews, can be
mined to devise marketing and advertising strategies. A Forrester research recently evaluated more
than 4,000 product reviews belonging to the Electronics and Home & Garden categories of Amazon.com and found that (a) 76 percent of people use customer reviews for shopping, (b) 84 percent
of product reviews were positive or neutral and, (c) despite negative product reviews
consumers understood that not all products or experiences are perfect.25
Many companies have now included a user review web page in their websites. For instance,
when CompUSA added a user generated review page along with the product search option, they
saw conversion rates rise 60 percent and individual spend rising 50 percent. Similarly, Yahoo! estimates that 40 percent of the online buying population is a product advocate and have an enormous
influence on buying habits and online communications patterns. Further, Bazaarvoice, a company
that provides online brands with managed solutions to enable, encourage and monitor online customer ratings and reviews found that nearly 77 percent of all customers use reviews and 81 percent
of shoppers who spend more than $500 online each month use product reviews when making buying decisions.26
Until recently, consumers were only talking amongst a small group of people. But with the
emergence of social media, consumers now have a wider and a more receptive audience. By listening to such responses about product and user experiences, companies can generate a two-way
conversation with the consumers and act on the reviews received. Such an exercise will boost loyalty among consumers.
Secondary Research
The main forte of the Internet is probably its advantages in researching secondary information. As
the 2005 Pew Internet & American Life Project pointed out, 73 percent of online adults utilize the
Internet for searching or gathering information. Another 67 percent use it for making a purchase.
The Internet competes with several other online resources to satisfy the information needs
of businesses and consumers. Businesses typically rely on professional databases such as
Lexis/Nexis or Knight-Ridder. Consumers most often utilize commercial online services such as
America Online or CompuServe. Although the latter also provide Internet access, they represent
a comprehensive information resource in themselves. Table 7-6 depicts how the Internet compares against other online information sources.
TABLE 7-6
Comparison of Information Sources
a
Internet
(WWW, Usenet)
Professional
Databases
(e.g., Lexis/Nexis)
Commercial
Online Services
(e.g., CompuServe)
Speeda
Low to high
High
Low to high
Information structure
Structure not very good
Very well structured
Well structured
Information scope
Very broad
Depends on the database,
mostly narrow
Medium to broad
Overall information quality
Low to high
High to very high
Medium to high
Search tools
Limited, not complete
Extensive, accurate
Mostly accurate
User support
Limited
Very good
Good
Cost
Low
High to very high
Medium
Depending on modem, server, and database structure.
163
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
The main advantages of the Internet at this point are its very broad scope, covering virtually every topic, and its comparably low cost. These characteristics make it a very appealing
medium to use for both consumers and businesses. Also, it is expected that the technical constraints of the Internet, such as low bandwidth, will be gradually resolved in the future, making
it less cumbersome to use in terms of downloading times.
Finding out information about competitor activities is an important task for businesses. The
Internet is a prime tool for this task, since it reduces the time spent and may increase substantially the quality of the information collected. Both product and financial information are probably suited best for competitive tracking. In particular, larger corporations display this information
most often in their websites. On the other hand, pricing information might not be amenable to
tracking readily, since it is not too common for businesses to display product prices (unless they
actually sell over the Internet). Competitive promotion and distribution information is probably
the least suited to tracking via the Internet. Information about products or companies can be
obtained using search engines on the Web. However, search engines have certain limitations and
hence do not guarantee that all relevant information has been obtained. For this purpose, there
are providers of custom search services, who search for information for a fee.
Custom Search Service
A growing number of custom search services are becoming available that offer information tracking and forwarding for a fee. These services range from one-time custom searches to regular
news deliveries. It is very common for these services to specialize in a particular area such as
global telecommunication or food processing and then offer their search results in a regular electronic newsletter to subscribers. Depending on the service, the charges can be based on usage
(one time or for the whole project), time frame (weekly, monthly, yearly), object (document fee,
fee per researched publication), amount of text (number of characters), or any combination of
these. For example, Dow Jones News/Retrieval Service provides customized business and financial news and information. Its service scans more than 6,000 business publications from around
the world, and the user can search 55 million business articles. The user retrieves the headlines
of the articles for free and pays $2.95 for each article viewed. Also, the customer can select “topic
folders” from a list of 1,200 preconfigured folders tracking information relevant to a user’s particular needs.
Agents
The use of intelligent agents in monitoring information is increasing rapidly. For illustrative purposes, let us consider a competitive analyst for a TV producer. The analyst needs constant information on the pricing of competitors’ products. An intelligent agent would monitor websites of
TV manufacturers and TV retailers, collect price and availability information, and deliver easyto-skim summaries.
Agents function in many different ways. Some are available on the Internet and preprogrammed for particular search tasks, such as Bargain Finders. This commercial application will
most likely gain tremendous significance in the near future. Other types of agents are a piece of
software that is purchased by the user, who then has to specify downloading criteria. Most often
searches are scheduled at night so they do not have to wait to download documents. Agent technology is on the verge of entering the mainstream of Web search, and the future is likely to bring
many innovations in this area.
Leveraging Web Analytics for Marketing
164
With the ever-changing dynamics of the World Wide Web, the relationship between consumers
and brands continues to evolve. Consumers are increasingly seeking avenues online to gather
THE INTERNET
US Dollars (in billions)
70
$57.5
60
MARKETING RESEARCH TODAY
$62.0
$52.8
$46.5
50
40
AND
$39.5
$32.0
30
20
10
0
2011
2012
2013
2014
2015
2016
FIGURE 7-1
U.S. Online Ad Spend 2011–2016
information about brands and to fulfill their need for products and services. As more consumers
continue to navigate online to connect with brands and to make purchase decisions, it is only
natural that marketers will seek to target them in this online space where time and convenience
are on the side of the consumer.
For the first time in U.S. history, marketers are projected to spend more on online advertising
than on advertising in print magazines and newspapers. As shown in Figure 7-1, online advertising is expected to generate $39.5 billion in sales in 2012—a 23.3 percent increase from 2011—compared to a sum of $33.8 billion on print. Online ad revenues are expected to continue to grow over
the next half-decade with total online ad investment projected to hit $62 billion by 2016.27
With online advertising increasing exponentially, marketers are interested in utilizing
resources to measure the effectiveness of their online communications in reaching consumers and
encouraging interaction with their brands. One such resource used to capture the online behavior of consumers and how people traverse through a site’s content is Google Analytics.
What is Google Analytics? Google Analytics is a free website statistics solution that gives rich insights
into web traffic and marketing effectiveness. It is the most widely used website statistics service with
60 percent of the 10,000 most popular websites using the service.28 It allows businesses to see a) how
people find their site, b) how they navigate through that site, and c) how they ultimately become customers. The tool allows marketers to monitor sales, conversion rates, and site engagement by generating detailed statistics about the visitors to a website. This tool can also track site visitors from all sources
including search engines, advertising, emails, and digital links.
Google Analytics offers tracking of campaign elements such as email, banner ads, and offline
ads to measure reach and effectiveness. The service is integrated with Adwords, an advertising
product that offers pay-per-click (PPC) advertising, cost-per-thousand (CPM) advertising, and
site-targeted advertising for text, banner, and rich-media ads. Subscribing to Adwords allows
businesses to display their ads on Google and its advertising network. Google Analytics can be
used to optimize the performance of Adwords by providing post-click data on keywords, search
queries, match type, and other ad components.
In terms of ecommerce, Google Analytics tracks sales activity and performance from products and
services sold online. The tool can trace transactions to campaigns and keywords to show which site content is generating the most revenue. In addition, the tool displays loyalty and latency metrics for purchases on the site. Purchase latency, the number of days between purchase events, can be monitored
to increase customer retention and revenues. Google Analytics can create a detailed overview of how
often customers in different segments are using the site’s products and services.
Additionally, the tool can also be used to track user interaction with websites and applications accessed from a mobile device. Traffic coming from smart phones can be tracked by device
or carrier and usage of apps can be tracked in the same way as a website.
How do you access it? Access to Google Analytics is easily obtained by signing-up for an
account and adding custom Java script provided by Google to each website page to be tracked.
165
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
FIGURE 7-2
Google Analytics Dashboard
Known as tracking code, the code collects visitor data and sends it to the Google server. The
data is then compiled by Google into a Dashboard with charts and graphs that display an
overview of the site’s performance. Figure 7-2 provides a screenshot of a typical Google Analytics Dashboard. The dashboard is customizable to user specifications. When users login to their
account, they are able to view visitor activity based on traffic, content, usage, location, ecommerce and other parameters selected by the user.
How can it be used for marketing purposes? Google Analytics makes it possible for advertisers, publishers, and website owners to improve their online results. Through this tool, marketers
can design targeted ads and strengthen their online marketing activities (such as online campaigns etc). Marketing Research in Action 7-3 shows how Costco Travel used Google Analytics to
increase its bottom-line. Some of the marketing implications of using such a tool are described
below.
Targeting consumers: By tracking the source of web traffic (such as search engine ads, email, referrals, etc.), marketers can better target consumers with marketing messages and ad placement that is
customized accordingly. This would eliminate the all too common mass advertising in the online
space. One can also create custom segments based on the most profitable groups of customers who
visit the website.
166
THE INTERNET
AND
MARKETING RESEARCH TODAY
MARKETING RESEARCH IN ACTION 7-3
Google Analytics in Action: Real World Implementation
Costco Travel is a wholly owned subsidiary of Costco Wholesale Corporation, the third largest retailer
in the U.S. with warehouses in seven other countries. Since August 2000, Costco Travel has been providing Costco members with vacation packages, cruises, hotels and car rentals. In September 2008, they
launched their online booking engine allowing members to shop and purchase vacation packages over
the Internet. In order to maximize performance of the travel site, Costco Travel evaluated multiple web
analytics products and chose Google Analytics in order to explore all possible options with the free
resource prior to seeking other paid services. The use of web analytics allowed Costco to analyze specific patterns in the behavior of their customers/visitors. Specifically, upon analyzing the landing page
and menu bar click patterns, Costco Travel was able to optimize page content which eventually resulted
in a large increase in traffic, c1ick-throughs and conversions. As a result, through Google Analytics,
online sales grew to over 50 percent of total vacation package sales and conversion rates increased. Thus,
Costco was able to leverage its online presence using the Google Analytics tool and thereby improve its
bottom-line to a great extent.
SOURCE: Adapted from case study, titled ”Costco Uses Google Analytics To Grow Costco Travel, Its Travel Booking Business”, (2010)
http://static.googleusercontent.com/external_content/untrusted_dlcp/www.google.com/en/us/analytics/case_studies/case_study_costco_en-US.pdf
Manage online advertising: With the ever increasing relevance of online advertising, marketers
(through web analytics tools such as Google Analytics) can make informed decisions on search
engine advertising decisions (through Google Adwords), webpage banner ad placement and online
campaign management. Valuable metrics such as revenue per click, ROI, and margin can be tracked
across various campaigns, thereby allowing marketers to manage online campaigns from the dashboard. Figure 7-3 is a screenshot of how managers can manage online campaigns on the Google
Analytics dashboard. Marketers can evaluate several campaigns simultaneously against several key
variables including visits, impressions, clicks, ROI and margin.
Understand regional distribution of visitors: In addition to understanding how visitors navigate
through the website, marketers (especially working with MNCs) would be interested to know
where their customers/visitors are located. Google Analytics also provides marketers with a mapping tool to describe the regional distribution of visitors (by continent, country, region, state &
city) to the website. This information can assist marketers in creating local search advertising or
offline ads/promotions in a specific region.
International Marketing Research
The Internet, as a global medium, increases the scope of communication tremendously. Even
though the globe-shrinking experience is already prevalent, the Internet’s ability to communicate
FIGURE 7-3
Campaign Tracking in Google Analytics
167
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
MARKETING RESEARCH IN ACTION 7-4
Attracting Online Customers with Samples
Putting up and promoting a site is one thing; getting prospective customers to visit is another. Aliments
E.D. Foods Inc., based in Pointe Claire, Quebec, Canada, has found an innovative way of achieving this.
Many people who shop online often look for free offers, and this is a good way to get them interested
in a product.
With a budget of just $500, Eiser, the founder of E.D. Foods, crafted a no-frills e-commerce website
that takes shoppers from the home page, to the products they want, to the transaction in as few clicks
as possible. Eiser decided that getting people to try their soup, gravies, and chilies was the key and
offered a free sample of the company’s soup mix to first-time buyers.
In 2001 and 2002, E.D. Foods sent out 30,000 free samples to first-time buyers. The promotion helped
E.D. Foods garner $500,000 in sales in 2001, and they expected that figure would jump to $750,000 by
March 2003. The average number of visitors at the site is about 1,500 a day.
According to Eiser, the Web is a funny place. It’s four times the amount of work you think it will be,
but when it works, it works really well.
SOURCE: “Samples Build Sales for Food Firm Online,” Marketing News, November 25, 2002.
across borders will have a dramatic impact. Marketing Research in Action 7-4 discusses how one
Web retailer has used a compelling tactic to capture new online consumers. Marketing Research in
Action 7-5 gives an illustration on the power of the Internet for international marketing research.
Particularly for businesses that are involved in international activities, the collection of intelligence from international resources will become natural. Also, databases such as STAT-USA are
available that are built and maintained by U.S. government agencies. These offer information to
companies entering the growing international trade market at no cost. Further, the most common Internet language is English. Figures 7-1 and 7-2 give you an idea of the way the World
Wide Web has penetrated the entire world and where the major users are located.
THE INTERNET AND MARKETING RESEARCH
DEVELOPMENTS
The time when one could keep up with the information on the World Wide Web is already
ancient history. With the Web growing dramatically, it becomes impossible to track even a small
and well-defined segment of the Web. Therefore, the market researcher has even more difficulty
finding the information he or she seeks. The purpose of this section is to point out some developments of the Internet which are most likely to happen in the near future and which will affect
the way a market researcher utilizes the Web.
Intranets
Intranets are internal company networks. While corporations are looking for ways and means of
communicating to consumers through the World Wide Web, it is apparent that intranets are the
building blocks for successful commercial activity. These internal networks start off as ways for
employees to connect to company information. Intranets may also incorporate connections to the
company’s various suppliers. According to many industry experts, the advent of total commercial integration is fairly close—employees, suppliers, and customers will soon operate in a totally
seamless environment. The advantage for an intranet user is that he or she can connect to the
Internet easily, whereas Internet users cannot access intranets without appropriate security codes.
The utilization of intranets will aid in the communication and distribution of information inside
large corporations. This is especially crucial for firms where information and know-how is missioncritical, such as management consultants or software developers. Once information is gathered, it is
168
THE INTERNET
AND
MARKETING RESEARCH DEVELOPMENTS
MARKETING RESEARCH IN ACTION 7-5
International Marketing Research on the Internet
Reeno, Inc., a chemical company based in Minneapolis, Minnesota, operates worldwide in the area of
agricultural inputs such as fertilizer, pesticides, and seeds. In the early 1990s the company developed a
new product line of agricultural seeds for improved cotton, rice, and corn varieties. After these products became an initial success in the home market, Reeno saw a tremendous potential for these products in countries such as India and China. These countries have a large and growing population, and
they have to satisfy a constantly growing demand for food. Therefore, making agricultural output more
efficient is a high priority in these countries.
However, Reeno knew from its past operations that introducing a new product in a country with a
completely different culture is not an easy task. They felt a need for more information regarding the purchase behavior and decision making of rural farmers in these countries. Reeno’s new line of seeds differed considerably from existing products, but the advantages of the new seed varieties had to be communicated properly. They appointed Innovative Marketing Consultants (IMC), a Houston-based
marketing research company, to do this research project. The task for IMC was to conduct secondary
research on the agricultural commodity markets in India and China. Also, IMC was to provide a report
on decision-making behavior of Indian and Chinese farmers.
The Internet was the natural first stop for Dr. Werner, the project supervisor at IMC. He started
with a very broad search utilizing a number of different search engines of the World Wide Web, such
as Alta Vista and Lycos. The search input was a number of different word combinations, such as
“farmer India,” “India cotton,” “corn production China,” or “rural China.” The use of
different search engines is important because no single engine indexes the whole WWW. These
searches led mostly to an extensive list of documents containing these search terms. Upon reviewing
the preliminary results and bookmarking the promising sites, Dr. Werner refined his search terms.
After a few hours he already had a fairly good idea regarding which websites contained relevant information. The next step was to examine the short list of promising websites. Here, the real advantages
of the Internet came into play, because many of the websites contained hyperlinks to related websites.
Because he was dealing with a narrow topic, he found himself several times linked to a website which
he had already visited: a true effect of the networked information structure. After half a day of intensive search he had a very good idea of the existing information resources on the Web with respect to
his topic, and he had collected an impressive volume of useful information. Among other things, he
obtained detailed information on agricultural crop harvests and production areas for both countries
for consecutive years from Cornell University in New York. Furthermore, the same resource supplied
information on seed and fertilizer imports and exports. In fact, he downloaded a Lotus 123 file that
contained all the data so that he could continue working with the file. The Internet archive of The
Hindu, India’s national newspaper (Madras), offered all farming-related stories within the last five
years. General country information about India and China was abundantly present. A rural sociology
discussion list from the University of Kentucky gave him a number of interesting facts about farmers’
practices in developing countries. From Lund University in Sweden he obtained an article written by
an agricultural sociologist about farming developments in India.
All in all, within less than a day Dr. Werner was able to gather an impressive amount of qualified
information on the subject. Even more impressive was the fact that he did this without even leaving his
office desk. He collected useful information in a fraction of the time he would have spent in libraries or
collecting other secondary resources. Also, he was able to pinpoint his search much more quickly than
he used to with printed resources.
stored in internal databases so that it can be accessed from any company location in the world. By
researching internal databases in the first place, the danger of duplicating information search procedures in separate locations is minimized and therefore the return on information is maximized. For
example, IBM maintains a large internal database to which every consultant worldwide has access.
When starting on a new project, the associate can research the database to find out if a similar project
has been worked on before (by IBM) and what types of information are already available.
Thus the Web can become an integral part of the information system of a corporation. “Firewalls” can protect from outside intrusion into the intranet. Sales, marketing, and communication
169
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
The Far-Reaching Web
Country
United States
Number of hosts
June 2011
Per 1000
inhab.
428,029,000
1156.8
Japan
63,199,811
495.6
Germany*
30,000,000
365.1
United Kingdom*
26,000,000
422.3
Italy
24,923,735
405.8
France*
24,900,000
385.0
Canada*
24,000,000
707.3
Brazil
22,212,190
116.0
China
17,446,117
13.0
Spain*
17,000,000
364.8
Australia
16,952,338
768.3
India*
16,000,000
13.8
Mexico
14,467,988
131.2
Netherlands
13,596,075
820.2
Russia
12,846,762
90.5
Poland
11,391,453
324.8
South Korea*
10,000,000
206.9
Argentina
8,628,736
215.0
Taiwan
6,405,566
277.1
Switzerland
5,232,713
672.3
Belgium
5,078,171
477.0
Sweden**
4,995,577
641.9
Finland
4,603,476
866.1
Denmark
4,242,576
766.5
South Africa
4,184,815
84.8
Turkey
4,034,284
55.6
Czech Republic
3,988,136
372.5
Portugal
3,558,417
332.3
Norway
3,542,038
729.1
Austria
3,389,616
405.7
Colombia
3,343,134
74.3
Greece
2,995,124
266.0
Thailand
2,980,010
44.0
Hungary
2,890,146
288.1
New Zealand
2,856,312
661.8
Romania
2,715,415
126.3
294.0
Israel
2,185,740
Ukraine
1,778,966
38.8
Chile
1,778,327
105.1
Singapore
1,701,103
341.0
Ireland
1,373,941
304.3
Slovakia
1,363,103
251.3
(continued)
FIGURE 7-4
SOURCE: http://www.gandalf.it/data/data1.htm.
*Figures for Germany, the UK, France, India, Canada, Spain and South Korea are “arbitrarily but not unreasonably” adjusted to compensate for underestimated hostcount data in this period.
**The number for Sweden includes hosts on “.nu” domains as well as “.se”.
170
ISSUES
Croatia
1,344,819
Indonesia
1,318,825
5.7
Lithuania
1,189,700
357.4
Bulgaria
909,615
120.0
Serbia
878,950
119.6
Hong Kong
845,018
120.7
Estonia
839,021
626.1
Uruguay
782,426
233.9
849,869,781
65.6
World total
AND
CONCERNS
301.5
Internet Users in the World
Distribution by World Regions—2011
12.0%
10.4%
6.2%
22.1%
3.4%
1.1%
Asia 44.8%
Europe 22.1%
North America 12.0%
Lat Am/Caribb 10.4%
Africa 6.2%
Middle East 3.4%
Oceania/Australia 1.1%
44.8%
FIGURE 7-4
(continued)
SOURCE: Internet World Stats, http://www.internetworldstats.com/stats.htm.
applications can combine internal and external information resources smoothly. Recently, networks between two companies have been set up to access information from both firms, and these
are termed as “Extranet.”
Speed
There are a number of promising technologies on the horizon which all have a common objective: to
increase the bandwidth of the Internet. The demand for high-speed connections is huge, since more
and more large data files such as multimedia applications are sent over the Internet. Backbone operators such as MCI and Sprint are starting to switch to fiber-optic networks which can accommodate
transmission of full-screen motion pictures. Marketing researchers will gain tremendously from this
development. For example, focus groups can be conducted from remote locations, and participants
who are thousands of miles apart can participate effectively. This will allow market researchers to discuss special topics with professionals or researchers who otherwise could not be brought together due
to time and travel constraints.
Work is currently being done in the area of photonic switching—switching light waves
without converting light to electrical energy. This work eventually will enable networks to take
the next leap to speeds not even dreamed of today.
ISSUES AND CONCERNS
Even as the use of the Internet gains rapid momentum worldwide, users and content providers
have to be aware of certain issues and concerns that go hand in hand with the various advantages of Web usage.
171
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
29
Identity Theft
As the Internet grows to reach a wider audience, it brings with it the problem of identity theft.
According to the Federal Trade Commission, identity theft occurs when a person obtains personal
information of others without permission, to commit fraud or other crimes. The type of information prone to theft includes details such as Social Security number (SSN), credit card number, and
driver’s license number among others pieces of identification. Victims are left with a tainted reputation and the complicated task of restoring their good names. The Javelin/Better Business Bureau
Report, released in January 2005, identifies that in 2004 nearly 9.3 million Americans were victims
of identity theft and the cost incurred due to identity theft is estimated at $52.6 billion.30 There are
four types of identity theft crime.31
1. Financial ID theft—Here the focus is on the victim’s name and Social Security number. The
imposter may apply for telephone service, credit cards, or loans, buy merchandise, and lease
cars or apartments.
2. Criminal ID theft—If stopped by law enforcement, the imposter in this case provides the victim’s information over his/her own. Eventually when the warrant for arrest is issued, it is in
the name of the person issued the citation—the victim.
3. Identity cloning—In this case the imposter uses the victim’s information to establish a new life.
Examples: Illegal aliens, criminals avoiding warrants, and people hiding from abusive situations or becoming a “new person” to leave behind a poor work and financial history.
4. Business or commercial identity theft—Businesses are also victims of identity theft. Typically the
perpetrator gets credit cards or checking accounts in the name of the business. The business
finds out when unhappy suppliers send collection notices or their business rating score is
affected.
In addition to conventional methods such as stealing wallets, dumpster diving, or theft of
paper mail, imposters also employ sophisticated means like accessing credit report information
by impersonating an employer, a family member, or a friend; getting hold of names and SSNs
from personnel files in the workplace, “shoulder surfing” in order to capture PIN numbers while
waiting at ATM machines and phone booths, and browsing the Internet to identify information
by locating public records sites and fee-based information broker sites. Some people even steal
personal information through e-mail or phone by posing as legitimate companies and claiming
that you have a problem with your account. This practice is known as “phishing” online, or
pretexting by phone. In fact, more than 20 percent of all cases involve telecommunications and
the Internet.
Thus, while several precautions are advised, the key to reduce the risk due to fraud lies in
closely guarding personal information, checking online accounts regularly to check for any fraudulent transactions, and monitoring credit history regularly. At the same time, online marketers
and researchers have to reassure customers and participants that their personal information will
be closely guarded and will not be misused in any way.
Privacy Issues32
172
In this digital era, the Internet has become an important part of daily life. An increasing number of
everyday activities are now being performed with the help of the Internet. This means that, given
the amount of data being exchanged every minute and the dearth of federal laws in the United States
governing disclosure of information to website visitors, the issue of online privacy is assuming great
significance.33 This affects individuals, companies, and societies as a whole. People would want to
maintain their privacy by controlling the information received about them by others. According to
Harris Interactive, the number of consumers unconcerned about privacy issues has decreased to 10
percent in 2003 from 22 percent in 1999.34
ISSUES
AND
CONCERNS
TABLE 7-7
Net Users Would Provide Personal Information For:
Incentive programs
76%
Members-only discounts
73%
Giveaways or sweepstakes
62%
Coupons
54%
SOURCE: Primary Knowledge/Greenfield Online. Adapted with permission from Michael Pastore, “Consumers Fear for Their Online Privacy,”
clickz.com, November 1, 1999, http://www.clickz.com/clickz/stats/1701261/consumers-fear-their-online-privacy
A major cause of concern in dealing with online privacy is about regulating the information
collected by companies that involve online transactions. The practice of placing files called
“cookies” and Web bugs by online companies in the user’s computer during a visit is quite common. These files capture information during an exchange between client and server. This trail of
bread crumbs may then be used to gather personal information about the user without their
knowledge. Typically, information such as time spent online, length of visits, type and name of
search engine accessed, products purchased, responses to promotions, and the navigation history are collected through these cookies.
While privacy issues are preventing some users from shopping online, more experienced
Internet users recognize data collection, and have embraced it in some respects. The “e-Customer
2000 Internet Consumer Survey” conducted by Primary Knowledge and Greenfield Online found
that online consumers are more accessible to basic data collection requests at websites, and are
increasingly willing to share personal information with e-marketers. The results from the study
are listed in Table 7-7.
Personalization is one of the reasons customers are willing to share their information
online. According to the Greenfield study, nearly 71 percent of the respondents have personalized a website, and two-thirds of that group believes personalization is important or very
important. Further, consumers are more likely to supply data such as birthday, household composition, or phone numbers when the data are used to create personalized content and pages.
Another reason for customer cooperation in sharing personal information is the customer
choice of contact. It was found out that when needed to contact, nearly 42 percent prefer to
contact the Internet retailer by e-mail and 36 percent prefer a toll-free number.
Thus, the response toward online privacy remains divided between users who refrain from
shopping online due to privacy concerns and those who do not mind providing their personal
information in return for benefits. This is reflected from a survey conducted by ChoiceStream in
2005, which reports that consumer willingness to allow sites to track clicks and purchases was
found to be 32 percent, down from 41 percent in 2004. This clearly indicates that while personalization is preferred by most consumers, the privacy fears associated with that continue to
grow.35 Between these conflicting groups of consumers are companies who are not really clear
about the privacy policies. Hence, to better address this issue, it is essential for companies to
develop and manage their customers’ privacy by taking a comprehensive and holistic outlook.
The Future of the Internet
With wireless connectivity so widespread, the Internet has become an integral part of daily life.
This is more so in the developed countries, where broadband adoption and broadband speed
provide instant access to the Internet. For example, there are more than a billion users in the
Asia/Pacific region, with an Internet penetration rate of over 50 percent as of March 2011. With
the rise of the Internet, an increasing number of services are gaining popularity.
Communication devices such as cell phones and PDAs are changing the face of Internetenabled devices with the ability to send and receive e-mail and access the Web. Through the
173
CHAPTER 7 MARKETING RESEARCH
ON THE INTERNET
widespread use of blogs, dramatic changes have happened to the media and publishing world.36
The Internet has also advanced the quality of human life by facilitating cutting edge technology. For instance, Electrolux, best known for its vacuum cleaners, has developed the ScreenFridge.37 It is an Internet icebox that manages the pantry, among other things. It e-mails a
shopping list to the local supermarket and coordinates a convenient delivery time according to
the schedule.
The Internet is thus one of the important sources of information. The speed, reliability, and
coverage of information are what make the Internet a fascinating medium. It has grown tremendously from a predominantly communication medium to a much larger and wider application. Its
presence can be felt from daily, routine activities to specialized research activities. The Internet also
has a flip side to it. The advancement of the Internet has given rise to Internet-based crimes like
identity thefts, privacy right violations, cyber stalking, and many more. As soon as one crime is
brought under control, another crime, bigger and much larger in size, comes up. Despite the contrasting sides to it, the Internet continues to provide a viable option for information for marketers
and consumers alike.
Summary
The Internet is a worldwide network of computers originally designed by the U.S. government to provide
an alternative communications network. Today, the Internet is a network of home and business users,
libraries, universities, organizations, and others. The Internet uses a common computer language, and there
are numerous communication services (for example, the World Wide Web) available for exploring the Internet. The Internet can level the playing field for marketing to potential customers through efficient use of
the medium. Both commercial and academic ventures have explored the characteristics of users of the Internet as well as uses for the Internet.
Collection of primary information on the Internet is increasing, as evidenced by the use of e-mail for
survey research. Other forms of primary data collection are performed primarily through interactive forms
which are filled out on the screen. Online focus groups are conducted entirely online—everything from
recruitment and screening to moderation of the discussion itself. Secondary information about products or
companies can be obtained using search engines on the Web. The power of the Internet for international
marketing research cannot be underestimated.
Questions and Problems
1.
2.
3.
John Smith has to make an unscheduled trip this
coming weekend to New York from his home town,
Houston. He searched on the Internet for the lowestprice offering and bought his ticket. Demonstrate the
search process that John used to get his ticket.
Assume that your family is interested in visiting the
NASA Space Center in Houston. They would like to
gather more facts about the place before they embark
on a trip. How can you assist them in this process?
Design a customer satisfaction questionnaire and
e-mail it to your friends. Ask them to respond to
your survey by filling out the questionnaire and
e-mailing it back to you.
174
4.
5.
You are hired as an intern in the marketing
research department of a chemical plant in the
United States. The firm is interested in expanding
its worldwide operations and is interested in
country-specific information regarding land use for
agriculture. How will you obtain this information
on the Internet?
Pick any company of your choice, and
a. Identify its website
b. Browse through the website
c. List the type of information that is available
through the website
End Notes
1. Although most of the Internet sites listed in this
chapter are current and updated, it is possible that
some of them might have been changed or linked
with another site. You should go through the search
procedure (for example, Net Search in Netscape) for
current information on relevant websites or visit
http://www.wiley.com for an index of updated links.
2. Enid Burns, “Internet Audience up 10 Percent
Worldwide,” March 6, 2007, http://www.clickz.
com/clickz/stats/1709218/internet-audiencepercent-worldwide, Retrieved on April 28, 2007.
3. Enid Burns, “E-Commerce Equals Convenience,
Risk to Consumers,” February 15, 2008, http://www.
clickz.com/clickz/stats/1705925/e-commerceequals-convenience-risk-consumers, Retrieved on
January 18, 2007.
4. Deborah Bulkeley, “Utah No. 1 in Homes with Computers,” Saturday, October 29, 2005, http://www.
deseretnews.com/article/635157158/Utah-No-1-inhomes-with-computers.html.
5. http://shop.org.
6. Jennifer Saranow, “Memo to Websites: Grow Up!,”
The Wall Street Journal, November 15, 2004, p. R.14.
7. Internet World Stats, http://www.internetworld
stats.com/stats.htm.
8. Robyn Greenspan, “Media Mixing for the Multicultural Market,” October 23, 2002, http://www.
clickz.com/clickz/stats/1707925/media-mixingmulticultural-market.
9. Forrester Research, http://shop.org.
10. Pew Internet & American Life Project, http://
shop.org.
11. Adapted with permission from Sean Michael
Kerner, “Online Ad Spend Swells,” April 1, 2005,
http://www.clickz.com/clickz/stats/1707269/
online-ad-spend-swells.
12. Fredrick Marckini, “Contextual Advertising,”
October 6, 2003, http://www.clickz.com/clickz/
stats/1715611/contextual-advertising-part.
13. Vauhini Vara, “Facebook Gets Personal With Ad Targeting Plan,” Wall Street Journal, August 23, 2007.
14. Don Bruzzone and Paul Shellenberg, “Track the
Effects of Advertising Better, Faster and Cheaper
Online,” July 2000, http://www.quirks.com/articles/
a2000/20000713.aspx?searchID=21438461&sort=
5&pg=1.
15. Rick E. Bruner, “The Decade in Online Advertising—
1994–2004,” April 2005, http://static.googleusercontent.
com/external_content/untrusted_dlcp/www.google
.com/en/us/doubleclick/pdfs/DoubleClick-042005-The-Decade-in-Online-Advertising.pdf.
16. Enid Burns, “Domain Registration Increased 32 Percent in One Year,” March 8, 2007, http://www.clickz.
com/clickz/stats/1710942/domain-registrationincreased-percent-one-year.
17. “Global Online Advertising Expenditures,” Marketing News, July 15, 2005, p. 28.
18. Donald R. Lehman and Russell S. Winer, Product
Management, Chicago, IL: Irwin, 1997.
19. Gabriel Gelb, “Online Options Change Biz a Little —
and a Lot,” Marketing News, November 1, 2006,
pp. 23–24.
20. J. H. Ellsworth and M. V. Ellsworth, The Internet
Business Book, New York: John Wiley, 1996.
21. Online Panels, Globalpark, http://www.globalpark.
com/market-research/online-panel.html.
22. Marketing Report, July 22, 1996, p. 4.
23. Allison Enright, “Real-time Analytics Boost ROI,
Accountability,” Marketing News, October 1, 2006,
pp. 20 & 24.
24. Steven Vonder Haar, “ROI Shows Marketers the
Money,” Inter@ctive Week, March 22, 1999, Vol. 6,
No. 12, p. 35.
25. Brian Kardon, “They’re Saying Nasty Things,”
Marketing News, December 15, 2007, pp. 30.
26. Emily Burg, “Leverage User-Generated Content
To Boost Brands,” http://www.mediapost.com/
publications/index.cfm?fa=Articles.showArticle
&art_aid=56933. Retrieved on September 17, 2008.
27. “Online spending will grow 23.3% to $39.5 billion
in 2012,” eMarketer, January 19, 2012, http://
www. emarketer.com/Article.aspx?id=1008783&R=
1008783.
28. “Google Analytics Usage Trends,” BuiltWith, February 12, 2012, http://trends.builtwith.com/analytics/
GoogleAnalytics.
29. http://www.ftc.gov.
30. “How Many Identity Theft Victims Are There?”,
September 2003, http://www.privacyrights.org/
ar/idtheftsurveys.htm.
31. http://www.idtheftcenter.org/
32. Adapted with permission from Michael Pastore,
“Consumers Fear for Their Online Privacy,” November 1999, http://www.clickz.com/clickz/stats/1701261/
consumers-fear-their-online-privacy.
33. Randall Frost, “Who Is Securing Your Identity
Online?” September 2003, http://www.brandchannel.
com/features_effect.asp?pf_id=177.
34. Harris Interactive, http://shop.org.
35. Sean Michael Kerner, “Consumers Want Personalization—and Privacy,” August 16, 2005, http://
www.clickz.com/clickz/stats/1716513/consumerswant-personalization-privacy.
36. Susannah Fox, Janna Quitney Anderson, and Lee
Rainie, “The Future of the Internet,” Washington,
D.C.: Pew Internet & American Life Project,
January 9, 2005.
37. http://www.learnthenet.com/index.php/html/
03future.htm.
175
w
ge / aak
w. wi l ey
om/ c o l
le
.c
CASE 7-1
Caring Children’s Hospital
er
w
Mary Beth, President of Caring Children’s Hospitals
(CCH), always looked forward to New Years. New
Years were meant for family reunions and holidays for
her. But this New Year had something totally different
in store for her. The previous year had not been good
for the hospital revenues. The number of patients
served by the hospital had gone down, and the costs of
health care were rising. The balance sheet presented a
very sorry picture to her. But there was a glimmer of
hope in the gloomy balance sheet.
About two years ago, Mary had started a small
division within CCH that provided health care for disabled children in their houses. This division resembled
a home health-care facility, but Mary was wary about
developing this division into a full-fledged home
health-care organization within CCH. This division had
been doing well for the past two years. It was the only
division in CCH with a positive picture in the balance
sheet. Mary decided to develop the home health-care
division in the hospital to get around the problems facing the hospital’s financial situation.
But there were a lot of issues to be addressed, such
as the kind of expansion plan to adopt, the kind of services to be provided by the home health-care facility, the
kind of insurance plans that the patients would be able
to use, and the type of publicity to use. Mary decided
to use the services of Innovative Marketing Consultants
(IMC), a top-notch local marketing consulting firm, for
addressing these issues and suggesting an expansion
plan. A meeting was set up with Dr. K, president of
IMC, to discuss the issues involved and to explain the
details of a home health-care program. Dr. K suggested
using the Internet to analyze the scope for a new player
in the industry, the trends in the industry, the common
modes of payment available, the costs involved per
patient, the profit margins possible, and the kinds of
diseases requiring home health care. The Internet was
selected for secondary data analysis because of its speed
and availability of up-to-date information.
Dr. K used several search engines available in the
Internet, such as Lycos, Infoseek, and AltaVista, to get a
wide and varied coverage of the Internet. The information collected from the Internet suggested that the home
health-care industry was booming and that pediatric
home health care had a significant share in the market.
The results also suggested that given a choice, people
would prefer to receive the needed care in their homes.
Also, the cost involved in serving patients at their houses
was significantly less and it was fully covered by
Medicare. Home health care was also found to be
the fastest growing health-care segment in the nineties.
The diagnoses most often referred to home health care
were found to be favorable to CCH. These results
formed the base for conducting primary research in the
market. The use of the Internet allowed Dr. K to analyze
a wide range of issues within a short time frame.
Questions for Discussion
1.
2.
3.
Use the Internet to corroborate the findings of Dr. K.
Prepare a report based on the information that you
have collected.
What are your recommendations to Mary, based on
your findings?
176
The case was prepared by V. Kumar and Rajkumar
Venkatesan for the purpose of classroom discussion.
INFORMATION COLLECTION:
QUALITATIVE AND
OBSERVATIONAL METHODS
8
Learning Objectives
■ Explain the need for qualitative research.
■ Introduce the different types of qualitative research methods.
■ Discuss in-depth interviews, focus-group, and projective techniques in detail.
■ Be familiar with the various observational methods.
This chapter shifts the focus from the utilization of already-available secondary data to the collection of primary data for a specific purpose. Seldom is enough known about a marketing problem or situation for the researcher to be able to proceed directly to the design of a structured
study that would yield representative and quantifiable results. The type of research one should
do is driven by the information needed. Marketing research can use either primary or secondary data, and it can also be either qualitative or quantitative.
Qualitative data collection is done to obtain a basic feel for the problem before proceeding
to the more analytical portion of the study.
A variety of qualitative methods can be used for such exploratory purposes. Specifically, we
will discuss individual and group interviews and case studies. The category of qualitative methods includes projective techniques that are used when self-reports are likely to be misleading.
Although projective techniques are utilized during exploratory research, they are also used for
primary data collection as well.
Observational methods are also discussed in this chapter. The observation of ongoing behavior is a widely used exploratory method, as well as an effective way to collect quantitative information when direct questioning is not possible.
NEED FOR QUALITATIVE RESEARCH
The purpose of qualitative research is to find out what is in a consumer’s mind. It is done in
order to access and also get a rough idea about the person’s perspective. It helps the researcher
to become oriented to the range and complexity of consumer activity and concerns. Qualitative
data are collected to know more about things that cannot be directly observed and measured.
Feelings, thoughts, intentions, and behavior that took place in the past are a few examples of
those things that can be obtained only through qualitative data collection methods. It is also
used to identify likely methodological problems in the study, and to clarify certain issues that
were not clear in the problem. Sometimes it may not be possible or desirable to obtain information from respondents by using fully structured or formal methods. Qualitative data collection methods are used in such situations. People may be unwilling to answer some questions
when confronted with them directly. Questions that they perceive as invasion of privacy, that
177
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
they think will embarrass them, or that may have a negative impact on their ego or status will
not be answered. Examples of such sensitive questions could be: “Are you a compulsive drinker
of alcohol? Do you use drugs to relieve stress or anxiety?” Sometimes, accurate answers will not
be forthcoming because they are part of the subconscious mind and cannot be tapped into
directly. They are disguised from the outer world through the mechanism of ego defenses, such
as rationalization. For example, a person may have purchased an expensive sports car to overcome a feeling of inferiority. However, if asked, “Why did you purchase this sports car?” he or
she may say, “I got a great deal,” or “My old car was falling apart.” It has been shown that
information of this sort can be better obtained from qualitative methods, such as focus-group
discussions or projective techniques, than through a formal, structured-survey method of data
collection.
The basic assumption behind qualitative methods is that an individual’s organization of a
relatively unstructured stimulus indicates the person’s basic perceptions of the phenomenon
and his or her reaction to it.1 The more unstructured and ambiguous a stimulus is, the more
subjects can and will project their emotions, needs, motives, attitudes, and values. The structure of a stimulus is the degree of choice available to the subject. A highly structured stimulus
leaves very little choice: The subject has unambiguous choice among clear alternatives. A stimulus of low structure has a wide range of alternative choices. If it is ambiguous, the subjects
can “choose” their own interpretations.
QUALITATIVE RESEARCH METHODS
Collectively, these methods are less structured and more intensive than standardized
questionnaire-based interviews. There is a longer, more flexible relationship with the respondent,
so the resulting data have more depth and greater richness of context—which also means a
greater potential for new insights and perspectives. The numbers of respondents are small and
only partially representative of any target population, making them preludes to, but not substitutes for, carefully structured, large-scale field studies. There are three major categories of acceptable uses of qualitative research methods:
1. Exploratory
■
■
■
■
■
Defining problems in more detail.
Suggesting hypotheses to be tested in subsequent research.
Generating new product or service concepts, problem solutions, lists of product features, and so forth.
Getting preliminary reactions to new product concepts.
Pretesting structured questionnaires.
2. Orientation
■ Learning the consumer’s vantage point and vocabulary.
■ Educating the researcher to an unfamiliar environment: needs, satisfactions, usage situations, and
problems.
3. Clinical
■ Gaining insights into topics that otherwise might be impossible to pursue with structured research
methods.
The range of possible applications of these methods can be seen from the following examples:
178
1. A telephone equipment supplier wanted to know what features to incorporate in an answering device located in a telephone substation (rather than in the home or office). From several group discussions came ideas for many features such as variable-length messages and
accessibility from any telephone. Specific features and price expectations were tested in a
subsequent survey.
2. In 2002, the photovoltaic (PV) solar electric system was installed in the state of Oregon. The
PV technology uses the sun’s energy to make electricity, unlike a solar hot water system
QUALITATIVE RESEARCH METHODS
which uses the sun’s energy to heat water. The Oregon Office of Energy was interested to
find out about the issues concerned with acquiring a PV resource. For this purpose, individuals who had been directly involved with the PV industry in Oregon were chosen to participate in focus groups.
The focus groups concentrated on the delivery, education, and installation of the PV solar
electric system in the state. In particular, issues relating to the major barriers as perceived by
the industry, the market opportunities, and the requirements to develop a grid-tied PV market
in Oregon were discussed. Results from the focus groups revealed the significant difference in
opinions about market development for PV systems. The complaints and comments relating
to the types of incentives, equipment performance, standards for equipment, and installer guarantees were then used in subsequent decisions such as consumer education, training, and easing the buying process of PV systems for the consumer.2
3. General Motors uses consumer and dealer focus groups,3 as well as extensive questionnaires, to
identify the best features of their own and of competitors’ automobiles. These give insight on
“world-class” elements they want to meet or exceed. For example, the design of air-filter covers for the GM10 line of cars was inspired by Mazda; the seat adjustment was made by a long
bar under each front seat; and electrical fuses were put in the glove box—an idea used by Saab.
Among the heaviest users of qualitative data are Japanese firms. They prefer “soft data” collected by managers during visits to dealers and customers, because it gives them a much better
feel for a market’s nuances. Talks with dealers who know their customers result in realistic, context-specific information. These talks relate directly to consumer attitudes, or the way the product has been or will be used, rather than being remote from actual behavior.4 If Japanese firms
do conduct surveys, they interview only people who have actually experienced the product or
service, rather than asking a random sample about general attitudes. When Toyota wanted to
learn what Americans preferred in small, imported cars, they asked groups of Volkswagen
Beetle owners what they liked and disliked about that particular car.
Use of Computers in Qualitative Research
Four major perceived constraints have traditionally mitigated the use of qualitative research:
1.
2.
3.
4.
Volume of data
Complexity of analysis
Detail of clarification record
Time-consuming nature of the clerical efforts required
Computer technology has helped alleviate most of these problems and has helped to increase
the use of qualitative research. The role of computer applications in the various tasks of qualitative data analysis is briefly described as follows:
Transcribing—Data has to be transcribed into text. A popular program used for qualitative analysis is Hyper TRANSCRIBE (www.researchware.com).
Storing—With the use of computer data, storage becomes very easy. It also brings more
sophistication in the organization of data. The researcher can access the material quickly
and precisely.
Coding—Codes have to be assigned by the researcher. Codes are labels assigned to data
segments to enable taxonomic organization of the data. Once codes have been
assigned, the computer can recall and print out all material belonging to a specific
code. Each extracted piece is logged with its source so that the researcher can easily
see from where it was extracted. The software can be used for complex, multiple coding. Coding systems can be reused much more easily and researchers can collaborate
directly with each other by working in parallel on the data analysis of one text, as
some programs (for example, Atlas/Ti and NVivo9 (www.qsrinternational.com) support the concept of multiauthoring for code systems.
179
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
Searching and retrieving—Researchers use not only standard Boolean operations (and, or, not),
but also various kinds of more complex search requests. For example, it is possible to
request the computer to perform a search such as “produce a list of quotations relating
to code ‘acceleration of a car’ as described by single men.”
Building relationships—The computer can be used to assist in the complex tasks of building
relationships between the data segments as part of the theory development process. In
some programs (for example, Atlas/Ti, Infination, and NVivo9), the relationship between
data segments can then be visualized as a network with annotated lines linking codes,
memos, or quotations.
Matrix building—Matrixes are cross-tabulation of variables, codes, or dimensions of the data
which help the researcher see how certain elements of the data interact. Many programs
(for example, XSIGHT from qsrinternational) feature a special matrix-building function
that speeds up this task.
Another perceived constraint slowing down the use of qualitative research is the comparatively higher expenses associated with the collection of qualitative data.
Individual In-Depth Interviews
Individual in-depth interviews are interviews that are conducted face to face with the respondent, in which the subject matter of the interview is explored in detail. There are two basic types
of in-depth interviews. They are nondirective and semistructured, and their differences lie in the
amount of guidance the interviewer provides. Marketing Research in Action 8-1 explores the reasons why one-on-one in-depth interviews should be considered more often by researchers.
MARKETING RESEARCH IN ACTION 8-1
Take a Qualitative Approach to Qualitative Research
In-depth interviews are a method that provides more information for less money without many of the
limitations focus groups can impose. One-on-one in-depth interviews should be considered more often
by researchers for a number of critical reasons:
1. More quality
Unlike in focus groups, one-on-one interviews can avoid responses that are influenced by other people. Interviewers may ask respondents directly and find out their personal thoughts on the product;
this enhances the quality of information.
2. More quantity
Researchers can receive twice the amount of information per respondent in an in-depth interview,
where the interviewer speaks at most 20 percent of the time, as in a typical 10-member focus group.
3. More depth
In-depth interviews capture all the relevance and salience of the qualitative information of focus
groups. Every word the respondent speaks can be taped and transcribed and used in multiple ways.
Well-designed surveys can go beyond surface answers and produce a rich database of interviews to
produce analyst reports, identify broad themes, and understand the ranges and depths of reactions.
4. More representation
In-depth interviews allow a much more representative approach as respondents are carefully selected
to represent the marketplace as accurately as possible.
5. More efficiency
Participants may be interviewed via a 15- to 45-minute phone conversation. Incentives of food and
money used in focus groups are not necessary for in-depth interviews.
6. More value
One-on-one interviews can double or triple the number of minutes that the respondent is talking, and
that is the true goal of research: understanding your consumers better.
For more interesting articles on marketing research visit http://www.marketingpower.com
180
SOURCE: Mark B. Palmerino, “Take a Qualitative Approach to Qualitative Research,” Marketing News, June 7, 1997.
QUALITATIVE RESEARCH METHODS
Nondirective Interviews. In nondirective interviews the respondent is given maximum freedom
to respond, within the bounds of topics of interest to the interviewer. Success depends on
(1) establishing a relaxed and sympathetic relationship; (2) the ability to probe in order to clarify and elaborate on interesting responses, without biasing the content of the responses; and
(3) the skill of guiding the discussion back to the topic outline when digressions are unfruitful,
and (4) always pursuing reasons behind the comments and answers. Such sessions normally are
one to two hours long and may be tape recorded (always with the permission of the respondent)
for later interpretation.
Semistructured or Focused Individual Interviews. In semistructured or focused individual
interviews the interviewer attempts to cover a specific list of topics or subareas, and the respondent has less freedom to extend the bounds of the topics. The timing, exact wording, and time
allocated to each question area are left to the interviewer’s discretion.
This mode of interviewing is especially effective with busy executives, technical experts, and
thought leaders. Basic market intelligence, such as trends in technology, market demand, legislation, competitive activity, and similar information are amenable to such interviews. The open
structure ensures that unexpected facts or attitudes can be pursued easily.
This type of interview is extremely demanding, and much depends on the interviewer’s skill.
First, the interviewer must be sufficiently persuasive to get through the shield of secretaries and
receptionists around many executives, in order to get an appointment. The major challenge is to
establish rapport and credibility in the early moments of the interview, and then maintain that
atmosphere. For this, there is no substitute for an informed, authoritative person who can relate
to respondents on their own terms. This can be achieved by asking the respondent to react to
specific information provided by the interviewer. Care should be taken to avoid threatening questions. A good opener might be, “If you had to pick one critical problem affecting your industry,
what would it be?” Cooperation sometimes can be improved by offering a quid pro quo, such as
a summary of some of the study findings.
A difficult problem with these interviews is the matter of record keeping. Some executives
dislike tape recorders, so it may be necessary to use a team of interviewers who alternate
between asking questions and note taking. To keep the interview as short as possible, it is usually best to leave behind a structured questionnaire for any specific data that are wanted, and
this can sometimes be assigned to staff for answering. Finally, since the appropriate respondents
for these studies are often difficult to identify, and may represent many parts of an organization, it is always advisable to ask for recommendations about which other people it might be
useful to interview.
Individual in-depth interviews are also used in consumer markets to identify key product benefits and trigger creative insights. Three techniques are being widely used now. In the
first technique, laddering,5 questioning progresses from product characteristics to user characteristics. A good starting point is with a repertory (sometimes called Kelly’s triad). If the
topic were airlines, respondents might be asked to compare one airline in a set of three to
the other two: How do airlines A and B differ from C? How do A and C differ from B? And
so on. Each attribute, such as “a softer seat,” is then probed to see why it is important to the
respondent; then that reason is probed, and so on. The result might be the following kind of
dialogue:
Interviewer:
Respondent:
Interviewer:
Respondent:
Interviewer:
Respondent:
“Why do you like wide bodies?”
“They’re more comfortable.”
“Why is that important?”
“I can accomplish more.”
“Why is that important?”
“I will feel good about myself.”
Notice that the dialogue has moved from a very tangible aspect of an airline to its contribution
to self-esteem.
181
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
The second technique is called hidden-issue questioning. In hidden-issue questioning, the
focus is not on socially shared values but rather on personal “sore spots”—not on general
lifestyles but on deeply felt personal concerns. The third technique, symbolic analysis, attempts to
analyze the symbolic meaning of objects by comparing them with their opposites. For example,
the following question could be asked: “What would it be like if you could no longer use airlines?” Responses such as “Without planes, I would have to rely on long-distance calls and letters” may be received. This would suggest that one of the attributes that can be highlighted in
an ad campaign for an airline could be face-to-face communication. Sometimes an interviewer
may have to go outside his country to interview people or may have to design questionnaires
that are to be administered in other parts of the world. Issues that an interviewer commonly
faces when interviewing in another culture are described in Marketing Research in Action 8-2.6
MARKETING RESEARCH IN ACTION 8-2
Qualitative Interviewing in the International Context
Evaluation has become an international activity. International and cross-cultural short-term evaluation site visits are much more subject to misinterpretations and miscommunications than traditional,
long-term anthropological fieldwork. The data from interviews are words. It is tricky enough to be
sure what a person means when using a common language, but words can take on a very different
meaning in other cultures. In Sweden, I participated in an international conference discussing policy evaluations. The conference was conducted in English, but I was there for two days, much of
the time confused, before I came to understand that their use of the term policy corresponded to my
American use of the term program. I interpreted policies, from an American context, to be fairly general directives, often very difficult to evaluate because of their vagueness. In Sweden, however, policies are very specific programs.
The situation becomes more precarious when a translator or interpreter must be used because of language differences. Using an interpreter when conducting interviews is fraught with difficulty. Special
and very precise training of translators is critical. It is important that questions be asked precisely as
you want them asked, and that full and complete answers be translated. Interpreters often want to be
helpful by summarizing and explaining responses. This contaminates the interviewee’s actual response
with the interpreter’s explanation to such an extent that you can no longer be sure whose perceptions
you have—the interpreter’s or the interviewee’s.
There are also words and ideas that simply can’t be translated. People who regularly use the language come to know the unique cultural meaning of special terms, but they don’t translate well. One
of my favorites from the Caribbean is “liming.” It means something like hanging out, just being, doing
nothing—guilt free. In conducting interviews for a program evaluation, a number of participants said
they were just “liming” in the program. But that was not meant as a criticism. Liming is a highly desirable state of being, at least to participants. Funders might view the situation differently.
The high esteem in which science is held has made it culturally acceptable in Western countries to
conduct interviews on virtually any subject in the name of science. Such is not the case worldwide. Evaluation researchers cannot simply presume that they have the right to ask intrusive questions. Many
topics may be taboo. I have experienced cultures where it is simply inappropriate to ask questions to a
subordinate about a superordinate. Any number of topics may be taboo, or at least indelicate, for
strangers—family matters, political views, who owns what, how people came to be in certain positions,
and sources of income.
There are also different norms governing interactions. I remember with great embarrassment going
to an African village to interview the chief. The whole village was assembled. Following a brief welcoming ceremony, I asked if we could begin the interview. I expected a private, one-on-one interview.
He expected to perform in front of and involve the whole village. It took me a while to understand
this, during which I kept asking to go somewhere else so we could begin the interview. He did not
share my concern about preference for privacy. What I expected to be an individual interview soon
turned out to be a whole village focus-group interview! In many cultures it is a breach of etiquette for
an unknown man to ask to meet alone with a woman. Even a female interviewer may need the permission of a husband, brother, or a parent to interview a village woman.
182
SOURCE: Michael Quinn Patton, Qualitative Evaluation and Research Methods, 2nd ed., Thousand Oaks, CA: Sage, 1990.
QUALITATIVE RESEARCH METHODS
Telephone depth interviewing and on-line interviewing using software like Skype are starting to gain greater acceptance among consumers and has proven to be more beneficial than focus
groups. When information is needed from business decision makers, strategically designed and
well-executed phone interviews may well turn out to be more successful than focus groups. For
one thing, phone interviews involve less time and expenditure of money. Also, interacting with
customers on an individual basis is more likely to elicit more detailed and useful information
than in a focus-group setting. Of course, the catch lies in employing experienced and intelligent
interviewers who are well briefed about the company, research objectives, and terminology likely
to be encountered. These interviewers should also be intelligent enough to sustain the interest
of the interviewees and come up with probing, open-ended questions, at the same time taking
care to avoid espousing personal judgments. The interviewers should also be well trained to listen and record responses verbatim. Listening has been the hallmark of many successful product
managers and CEOs. The idea is to establish a personal rapport with interviewees that encourages them to talk freely and without inhibition. The most important step in phone interviews is
capturing the data in a software program that can code and tabulate responses from open-ended
questions. This step could also be done manually, but using suitable software enhances credibility by capturing the responses verbatim and also making them available for future use.7 More
telephoning is being used in qualitative research because of the following factors as well: The
telephone has become a standard medium of communications, clients are more receptive to saving time and therefore money, and studies can be conducted in remote areas that other qualitative methods, such as focus groups, cannot access.
Technology is chipping in by helping with the analysis of individual in-depth interviews.
Here’s how it works: A topic guide is entered into the computer in advance; the interviewer,
with a headset on for comfortable, hands-free interviewing, types answers directly into the program. Comments of any length can be accommodated, wherever they occur, and skip patterns
are easily handled. Afterward, a printout shows all responses to each open-ended question, either
for the full sample or by subgroup (male/female, age group, heavy users, former subscribers,
etc.). Individual interviews and interview summaries of each interview can also be printed out.
The program provides frequency counts and averages on the closed ends.8
Focus-Group Discussions
A focus-group discussion is the process of obtaining possible ideas or solutions to a marketing
problem from a small group of respondents by discussing it. The emphasis in this method is on
the results of group interaction when focused on a series of topics a discussion leader introduces.
Each participant in a group of five to nine or more persons is encouraged to express views in
each topic and to elaborate on or react to the views of the other participants. The objectives are
similar to unstructured in-depth interviews, but the moderator plays a more passive role than
an interviewer does.
In conducting focus groups, it is very important to position focus-group observers by spending time to brief them before the groups start and debrief them afterward. Positioning client
observers behind one-way viewing windows can avoid problems of misinterpretation of results
in the future, and prepare observers to gain a better understanding of what they will be seeing
and hearing. Observers are less prone to assume that the opinions of one or two participants are
somehow representative of any group of people other than the focus-group participants themselves. Some of the issues that need to be addressed in the briefing session are
1. Outlining the intended direction of the group
2. Explaining how participants were recruited
3. Reeducating observers on the concepts of random selection, statistical reliability, and projectability of research results.9
The focus-group discussion offers participants more stimulation than an interview; presumably this makes new ideas and meaningful comments more likely.10 Among other advantages, it
183
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
is claimed that discussions often provoke more spontaneity and candor than can be expected in
an interview. Some proponents feel that the security of being in a crowd encourages some participants to speak out. Marketing Research in Action 8-3 illustrates two scenarios of focus-group
discussion. The following photo shows a typical focus group room.
A focus group in progress with researchers observing the
group discussion through a one-way mirror.
SOURCE: © Marmaduke St. John/Alamy
Types of Focus Groups. Focus groups can be classified into three types. Exploratory focus
groups are commonly used at the exploratory phase of the market research process to aid in
defining the problem precisely. They can also be viewed as pilot testing: Exploratory groups can
be used to generate hypotheses for testing or concepts for future research.
Clinical focus groups involve qualitative research in its most scientific form. The research
is conducted as a scientific endeavor, based on the premise that a person’s true motivations
and feelings are subconscious in nature. The moderator probes under the level of the consumer’s consciousness. Obviously, clinical groups require a moderator with expertise in psychology and sociology. Their popularity is less because of the difficulty of validating findings
from clinical groups and because unskilled operators sometimes attempt to conduct clinical
groups.
The reality in the kitchen or supermarket differs drastically from that in most corporate offices.
Experiential focus groups allow the researcher to experience the emotional framework in which
the product is being used. Thus an experiencing approach represents an opportunity to
“experience” a consumer in a natural setting.
Carmen Sandoes, marketing manager for the Bronx Zoo, says the nonprofit institution used
focus groups to test various appeals to increase the number of zoo visitors. Participants were
divided evenly between those who had and had not been to the zoo. They were asked to pick
animal concepts—such as “Great Snakes Day” and “Big Bears Week”—which could be used as
special-events themes. These themes—promotional tools—would feature exotic snakes for a day
or display grizzly and polar bears for a week. Using the focus group helped Sandoes and her
staff to find which animals aroused the most interest and to implement their discovery in the
form of future special events.
There are no hard-and-fast rules for choosing focus groups rather than individual in-depth
interviews for qualitative studies. The comparison in Table 8-1 may help you to make the
choice.11
184
Key Factors for Focus-Group Success. As a rule, three or four group sessions usually are sufficient. The analyst invariably learns a great deal from the first discussion. The second interview
QUALITATIVE RESEARCH METHODS
MARKETING RESEARCH IN ACTION 8-3
A Tale of Two Focus Groups
It’s Wednesday evening, and the first focus group of the night at a New Jersey facility is winding up a
session paid for by a mid-sized Eastern city that wants nothing more than to become a bonafide tourist
destination. This group, composed entirely of people over 50 who already have visited the destination,
seems reluctant to have the session end. The city in question is an important site in African-American
history, and several well-informed black women are relishing their reminiscences of it. One very animated
woman seems pleased that she knows a number of facts that the moderator doesn’t.
She mentions that the city is also the site of an annual basketball playoff for African-American colleges.
A white man wonders whether people who are not African-American would hesitate to visit the city were
it to advertise its connection to that event. A heated discussion breaks out, and the other participants agree
that the man is off the mark.
After the moderator ends the session—which went 10 minutes past its 90-minute time slot—several
of the members of this well-informed group shake hands and say how much they’ve enjoyed meeting
each other.
Now it’s time for the 6:00–7:30 P.M. session. A decidedly tired-looking group of men and women, all
parents between the ages of 30 and 45, plop down in their chairs. Maybe it’s because the participants
haven’t had dinner yet, or miss their kids, or had tough days, or intimidate each other, or have lackluster personalities, but this crowd is much less expressive. They also don’t seem as well-informed.
Several seem as if they’d probably criticize just about anything on general principle.
A mother of two, friendly and eager to please, says she found the city’s history fascinating and she
believes others would, too. The man next to her—a naysayer if ever there were one—says he ended up in
the city by accident and found nothing much of interest there. Several minutes later, the woman unaccountably switches her mind and says that she doesn’t think the city’s history would interest many people. A heavy atmosphere settles over the room. Everyone looks drained and no one ventures a strong view.
Interestingly, given the differences in the two panels, both groups were chosen through random phone
surveys conducted by the facility personnel. The panelists were asked their ages, number of children,
and familiarity with the city. They also were asked if they had participated in similar sessions. In recent
years the industry has been plagued by “focus-group junkies” who show up repeatedly. Recently, annoyed
sponsors have begun demanding fresh blood.
Both panels were asked to give reactions to eight possible slogans for the city. Surprisingly, both
groups chose the same three slogans, each of which underscored the city’s diverse attractions, such as
restaurants, shopping, and children’s amusements. The participants’ consensus was that a broad-based
pitch would attract the largest number of tourists. Similarly, they rejected several slogans playing up the
city’s role in history, concluding that such a focus would be too narrow.
Client John Boatwright, who owns a Virginia tourism consulting group called Boatwright & Co., says
it is occurrences like very different groups arriving at identical results that give him confidence in focus
groups. Clients tend to conclude they are on the right track when such coincidences take place, he says.
“If you do four to six groups and get a consensus, you can conclude that if you did exponentially more
groups, you would get the same results,” he explains.
For more articles from Forecast Magazine visit http://adage.com/section/american-demographics/195.
SOURCE: Leslie Wines, “A Tale of Two Focus Groups,” Forecast Magazine, July/August 1995, p. 27.
produces much more, but less is new. Usually, by the third or fourth session, much of what is
said has been heard before, and there is little to be gained from additional focus groups. Exceptions to this rule occur if there are distinct segments to cover, such as regional differences in
tastes, the differences between women working in the home and outside the home, or the differences between married or unmarried women.
A focus group is not an easy technique to employ. Further, a poorly conducted or analyzed
focus group can yield very misleading results and waste a good deal of money.12 The recruitment costs, payments to participants, space rental, moderation, and analyst fees easily are in the
range of $5,000 to $7,500 per focus group for consumer studies.13 Ten tips for running successful focus groups are presented in Marketing Research in Action 8-4.
185
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
TABLE 8-1
Comparison of Focus Groups and Individual In-Depth Interviews
Focus Groups
Individual In-Depth Interviews
Group interactions
Group interaction is present. This may
stimulate new thoughts from
respondents.
There is no group interaction. Therefore,
stimulation for new ideas from
respondents comes from the interviewer.
Group/peer
pressure
Group pressure and stimulation may
clarify and challenge thinking.
In the absence of group pressure, the
thinking of respondents is not challenged.
Peer pressure and role playing may
occur and may be confusing to interpret.
With one respondent, role playing is
minimized and there is no peer pressure.
Respondent
competition
Respondents compete with one another
for time to talk. There is less time to
obtain in-depth details from each
participant.
The individual is alone with the
interviewer and can express thoughts in
a noncompetitive environment. There is
more time to obtain detailed
information.
Influence
Responses in a group may be
“contaminated” by opinions of other
group members.
With one respondent, there is no
potential for influence from other
respondents.
Subject sensitivity
If the subject is sensitive, respondents
may be hesitant to talk freely in the
presence of several other people.
If the subject is sensitive, respondents
may be more likely to talk.
Interviewer fatigue
One interviewer can easily conduct
several group sessions on one topic
without becoming fatigued or bored.
Interviewer fatigue and boredom are
problems when many individual interviews
are needed and they are conducted by one
interviewer.
Amount of
information
A relatively large amount of information
can be obtained in a short period
of time at relatively modest cost.
A large amount of information can be
obtained, but it takes time to obtain it and
to analyze the results. Thus, costs are
relatively high.
Stimuli
The volume of stimulus materials
that can be used is somewhat limited.
A fairly large amount of stimulus material can be used.
Interviewer
schedule
It may be difficult to assemble
8 or 10 respondents if they are a
difficult type to recruit (such as very
busy executives).
Individual interviews are easier to
schedule.
SOURCE: Adapted from Focus Groups: Issues and Approaches (New York: Advertising Research Foundation, 1985).
186
Planning the Agenda. Planning starts by translating the research purpose into a set of managerially relevant questions, which ensures that client and moderator agree on specific objectives before the study begins. From these questions, the group moderator can prepare a discussion guide to serve as a checklist of the specific issues and topics to be covered. However,
this list is strictly for general guidance; it is not desirable to read formal questions to the group.
An important issue is the order in which the moderator introduces topics. Usually, it is best
to proceed from a general discussion to increasingly specific questions, for if the specific issue is
addressed first it will influence the general discussion. It is also easier for respondents to relate
to a specific issue when it has been preceded by a general discussion. For example, Mother’s
Cookies was interested in concept-testing a new fruit-filled cookie, and a proposed introductory
promotion involved tickets to a circus performance. The moderator started with a general discussion about snacks and then moved to the use of cookies as snacks and the question of buying versus making cookies. Only after this general discussion was the more specific topic
addressed.
The set of topics covered may change after each focus-group experience. The moderator and
client may decide that a question is not generating useful, nonrepetitive information, and drop
QUALITATIVE RESEARCH METHODS
MARKETING RESEARCH IN ACTION 8-4
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Ten Tips for Running a Successful Focus Group
You can never do too much planning for a focus group. Developing the discussion guide with the
topics to be covered is critical.
Manage the recruitment process actively to be sure to get the right people in the groups.
Don’t prejudge the participants based on physical appearance.
The best focus-group moderators bring objectivity and expertise in the process to a project.
Achieving research objectives does not guarantee a successful focus-group project.
The moderator and the client should coordinate their efforts at all stages of the process for the
research to achieve its objectives.
Most client organizations conduct more focus groups than are necessary to achieve the research
objectives.
One of the most important services a moderator can provide is a fast report turnaround.
Client observers should be thoroughly briefed about the research objectives before the sessions start.
The most valuable service a moderator can provide is objective conclusions based on the interpretation
of the research, without regard for what the client wants to hear.
SOURCE: Thomas Greenbaum, “10 Tips for Running Successful Focus Groups,” Marketing News, September 14, 1998.
it from the remaining focus groups. Or a new, interesting idea may emerge, and reactions to that
idea may be sought from subsequent groups.14
Recruitment. When recruiting participants, it is necessary to provide for both similarity and contrast within a group. However, as a rule, it is undesirable to combine participants from different
social classes or stages in the life cycle, because of differences in their perceptions, experiences,
and verbal skills.15
Within an otherwise homogeneous group, it may be helpful to provide for a spark to be
struck occasionally, by introducing contrasting opinions. One way to do this is to include both
users and nonusers of the product or service or brand. If the product carries social connotations,
however, this mixing up may suppress divergent opinions; for example, buyers of large life insurance policies may believe that nonbuyers are irresponsible. Some moderators believe that having conflicting opinions within a group may invite either a “rational” defense or a “withdrawal”
of those who think their opinions are in a minority.
One controversial source of participants is the “experienced” panel, whose members have
been trained in ways that contribute to the dialogue in the group. Those who oppose this practice feel that “professional” respondents who show up repeatedly are so sensitized by the interview experience that they are no longer representative of the population.16
Although groups of 8 to 12 have become customary, smaller groups may be more productive.17 With 12 panelists, for example, after subtracting the time it takes to warm up (usually
about 3 minutes) and the time for the moderator’s questions and probes, the average panelist in
a 90-minute focus group has 3 minutes of actual talking time. The experience becomes more like
a group survey than an exploration of experiences, feelings, and beliefs. It is also a very expensive form of survey, so cutting group size makes sound economic sense.
The Moderator. Effective moderating encourages all participants to discuss their feelings, anxieties, and frustrations as well as the depth of their convictions on issues relevant to the topic,
without being biased or pressured by the situation.18 The following are critical moderating skills:
■ Ability to establish rapport quickly by listening carefully, demonstrating a gen-uine interest in
each participant’s views, dressing like the participants, and avoiding the use of jargon or sophisticated terminology that may turn off the group.
■ Flexibility, observed by implementing the interview agenda in a way the group finds comfortable. Slavish adherence to an agenda means the discussion loses spontaneity and degenerates into a question-and-answer session.
187
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
■ Ability to sense when a topic has been exhausted or is becoming threatening, and to know
which new topic to introduce to maintain a smooth flow in the discussion.
■ Ability to control group influences to avoid having a dominant individual or subgroup that
might suppress the total contribution.
Common techniques for conducting successful focus-group interviews include the chain reaction, devil’s advocate, and false termination. In the chain reaction technique, the moderator
builds a cumulative effect by encouraging each member of the focus group to comment on a
prior idea suggested by someone else in the group, by adding to or expanding on it. When playing devil’s advocate, the moderator expresses extreme viewpoints; this usually provokes reactions
from focus-group members and keeps the discussion moving forward in a lively manner. In false
termination, the moderator falsely concludes a focus-group interview, thanks group members for
participating, and inquires whether there are any final comments. These “final comments” frequently lead to new discussion avenues and often result in the most useful data obtained.
Analysis and Interpretation of the Results. Analysis and interpretation of the results is complicated by the wealth of disparate comments usually obtained, which means that any analyst
can find something that agrees with his or her view of the problem. A useful report of a group
session is one that captures the range of impressions and observations on each topic and interprets them in the light of possible hypotheses for further testing. When reporting comments, it
is not sufficient merely to repeat what was said without putting it into a context, so that the
implications are more evident.
Several features of group interactions must be kept in mind during the analysis. An evaluation of a new concept by a group tends to be conservative; that is, it favors ideas that are easy
to explain and not necessarily very new. There are further problems with the order of presentation when several concepts, products, or advertisements are being evaluated. If group participants have been highly critical of one thing, they may compensate by being uncritical of the next.
Marketing Research in Action 8-5 discusses Procter and Gamble’s experience with focus groups
while developing new products.
Trends in Focus Groups. A growing number of focus groups have migrated from the real to
the virtual world over the past few years. Given the cheaper costs, speed, and new technologies that will facilitate Web-based communication, the number of virtual focus groups will keep
increasing but will never entirely replace their real-world counterparts.19 Marketing Research
in Action 8-6 points out the advantages and disadvantages of online versus traditional focus
groups. The number of focus groups being conducted is growing at a rapid pace. There may
be over 50,000 focus groups conducted annually. The quality of focus groups’ facilities have also
become better. Instead of tiny viewing rooms with small oneway mirrors, plush two-tiered
observation areas that wrap around the conference room to provide an unobstructed view of
MARKETING RESEARCH IN ACTION 8-5
The Sweet Smell of Success
When Procter & Gamble Co. were researching to build a better air freshener, they decided that they
would ask the consumers about their “desired scent experience.” Through this exercise, researchers at
P&G found out that many people seemed to adjust to a scent after about half an hour and could not
smell it anymore. They also learnt that most air-freshener scents do not spread evenly across a room and
people felt many scents smell artificial.
After considering all these views, P&G developed Febreze Scentstories, a scent “player” that looks
like a CD player and plays one of five alternating scents every 30 minutes. Priced at $34.99, the player
contained a tiny fan inside that circulated the scent throughout the room. Along with this player, P&G
offered five different disks holding a variety of scents with trademarked names.
188
SOURCE: Deborah Ball, Sarah Ellison and Janet Adamy, “Just What You Need!,” The Wall Street Journal—Marketplace, October 28, 2004.
QUALITATIVE RESEARCH METHODS
MARKETING RESEARCH IN ACTION 8-6
Online Focus Group
An online focus group is very similar to a traditional focus group, but it takes place on the Web rather
than in a physical location. The sessions have a moderator, and the participants interact via chatlike
interfaces. With technology advancing by the day, researchers are increasingly attracted to online focus
groups. They started off just as modified chat rooms, but their capabilities have grown to include access
to multiple people, who can view the session as it unfolds. Viewers can even make comments to the
moderators without alerting participants. In short, online focus groups have been transformed into
sophisticated research environments. Online qualitative research is especially good for some of the following situations:
■ Studies in potentially sensitive or confidential areas where anonymity is essential
■ B2C studies with a widely dispersed (including rural areas) audience
■ Studies where it is not economically feasible to bring respondents together under the same roof
because of small numbers involved
■ Studies related to information technology that involve feedbacks on related topics and website
evaluations
■ Studies where professionals are involved and time is a constraint (for example, B2B customers)
In the situations mentioned above, online focus groups can perform as well as or better than traditional focus groups. But they may not always work best. The following are situations for which online
focus groups may not be the best approaches:
■ When capturing body language or facial expressions is vital
■ When you need to show prototypes or three-dimensional models
■ When products need to be handled—that is, when hands-on usage is critical or touching/feeling experience is mandatory
■ When conducting taste-testing, commercial testing, or testing of ads with extensive copy
■ When the client material or the topic is highly confidential
Most researchers who have used online focus groups agree that online respondents tend to be more
candid and direct. Generally, respondents compose their answers before reading others’ postings and
express only their own opinions, avoiding the peer pressure sometimes experienced in traditional groups.
Respondents also seem to express their opinions without the tempering sometimes seen in traditional
groups. For example, respondents feel more comfortable giving you negative or controversial feedback.
SOURCE: Monica Zinchiak, “Online Focus Group FAQs,” Quirk’s Marketing Research Review (www.quirks.com), July 2001.
all the respondents are being used. Telephone focus groups have emerged recently. This technique has been developed for respondents who are difficult to recruit, such as doctors. These
focus groups use the conference calling facility to conduct the discussion. Marketing Research
in Action 8-7 discusses three identifiable periods of qualitative research of women, as they have
become increasingly involved in focus groups over the decades.
In cities where focus-group facilities are not available, a focus group can be conducted in
church basements, in restaurants, and in a variety of hotel meeting rooms. One of the most successful groups was held in what the hotel called its boardroom. This was an elegant room containing a handsome conference table seating 12, with very plush executive-style chairs, deep pile
carpeting, and subdued lighting. Through the hotel, valet parking was arranged for the participants, as well as a gourmet meal, a cash honorarium, and (perhaps the most unexpected touch)
having all their cars lined up and waiting as the focus group ended. Client reps watched and listened to the group through a video feed to an adjoining meeting room. While this lacked some
of the immediacy and intimacy of viewing through a mirror, it still gave the client the sense of
being there. And the participants, the CFOs of area hospitals, enthusiastically responded with a
189
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
MARKETING RESEARCH IN ACTION 8-7
Focus on Women: Three Decades of Qualitative Research
There are three identifiable periods for qualitative research of women:
1. The early 1970s—the end of the traditional era
■ Focus groups were done during the day (assuming all women were housewives).
■ Women were paid less than men for attending: $7.50 vs. $10.00.
2. The late 1970s through late 1980s—the discovery of the new working woman
■ Market researchers started including women in studies for “nontraditional” products (for example, air travel, investments, cars).
3. The late 1980s to the present—the end-of-the-century woman
■ Researchers now routinely include women in what previously had been all-male groups.
■ Women hold their own mixed-gender focus groups on such nontraditional categories as technology, financial management, and sports.
For more interesting articles on marketing research visit http://www.marketingpower.com
SOURCE: Judith Langer, Marketing News, September 14, 1998, p. 21.
candid and lively discussion of the key issues.20 Marketing Research in Action 8-8 shows how
Internet technology serves the traditional focus group by reducing costs and facilitating greater
participation from people.
Recently, marketers have begun including creativity, in addition to the principles of marketing science, in the spectrum of focus groups. Creative moderating processes augment questioning, thus better engaging the group’s thinking and concentration and resulting in a more indepth and revealing feedback. Marketing Research in Action 8-9 lists three of the most effective
interactive exercises.
Another emerging trend is in two-way focus groups.21 This allows one target group to listen to and learn from a related group. In one application, physicians viewed a focus group of
arthritis patients discussing the treatment they desired. A focus group of these physicians was
then held to determine their reactions. A new focus-group television network called the Focus
Vision Network may represent a third trend. Instead of flying from city to city, clients can view
the focus groups in their offices. Live focus groups are broadcast by video transmission from a
nationwide network of independently owned focus facilities. The cost of this option, of course,
is quite high.22 Focus Vision also plans to unveil an international network of focus facilities. Thus,
global focus groups will be possible in the near future.23
MARKETING RESEARCH IN ACTION 8-8
Internet Technology Helps Traditional Focus Groups
Focus group, one of marketing’s favorite research tools, has now been made even more attractive and
user-friendly. Video-streaming technology now means that focus groups can be observed “live” from
one’s desk, while the focus-group technique itself remains unchanged. Participants gather at a location
to discuss and talk under the guidance of a group facilitator. However, observers no longer need to be
on-site, behind the mirror, to view the proceedings. A camera captures all the action close-up, including
facial expressions, and broadcasts the action via video streaming to an unlimited number of viewers who
can watch in real time from the comfort of their desktop computers at any time, in any place. Once the
focus-group session is complete, the data are saved to a server, where it can be viewed again by the
client on demand. Major benefits of the technology are reduced travel costs and an expanded viewing
audience.
190
SOURCE: Alf Nucifora, “Internet Revolutionizes Focus Groups,” American City Business Journal, September 2000.
QUALITATIVE RESEARCH METHODS
MARKETING RESEARCH IN ACTION 8-9
Effective Interactive Focus-Group Exercises
Focus-group interactive exercises are not only instrumental in eliciting in-depth feedback based on the
perceptions and emotions of respondents; they are also useful in improving their thinking and concentration. In addition, interactive exercises may uncover issues that respondents had been previously
unable to verbalize or had been unaware of. Such exercises must be in line with the research objectives
and should be relevant to the respondent base. The following are three effective interactive focus-group
exercises:
■ Product sort: This exercise presents the respondents with a sample of products that are to be sorted
into groups that make sense to them. They are then asked to give each group a name that describes
why those items have been put together. Thus, this exercise provides key insights into the segmentation process through visual simulation and active discussion.
■ Storytelling: Here the moderator describes a scenario (related to the product or brand) and then asks
the respondents to tell a story related to that setting. The process of weaving a story around the product or brand reveals the perceived cues, images, and biases of the respondents. This exercise is helpful in understanding the attitudes of the consumer.
■ Sticker allocation: This exercise provides the respondents with a short list of choices, such as product concepts or flavors. They are then asked to assign 10 stickers or purchase coupons according to
their preferences. Though the data from this exercise might not be statistically significant, the discussions during the exercise might throw more light onto customer preferences.
These dynamic interactive exercises not only stimulate lively discussions but also increase consumer
learning.
SOURCE: Holly M. O’Neill, “Interactive Exercises Better Engage Groups,” Marketing News, March 3, 2003, p. 55.
Using Projective Techniques in Qualitative Research
The central feature of all projective techniques is the presentation of an ambiguous, unstructured
object, activity, or person that a respondent is asked to interpret and explain.24 The more ambiguous the stimulus, the more respondents have to project themselves into the task, thereby
revealing hidden feelings and opinions. These techniques often are used in conjunction with
individual nondirective interviews.
Projective techniques are used when it is believed that respondents will not or cannot
respond meaningfully to direct questions about (1) the reasons for certain behaviors or attitudes,
or (2) what the act of buying, owning, or using a product or service means to them. People may
be unaware of their own feelings and opinions, unwilling to make admissions that reflect badly
on their self-image (in which case they will offer rationalizations or socially acceptable
responses), or are too polite to be critical to an interviewer.
Originally, projective techniques were used in conjunction with clinical “motivation research”
studies. One such study was done on Saran Wrap, a plastic food wrap, when it was first introduced. Because it was very clingy, it was effective in sealing food, but it was also quite difficult
to handle. As a result, strong negative attitudes toward the product became evident. To clarify
the reasons for this dislike, a series of in-depth, nondirective clinical interviews was conducted.
During the 1950s there were many homemakers who disliked or even hated their role of keeping house and cooking. At that time, prior to the resurgence of the women’s movement, there
was no acceptable outlet among women for this dislike. It could not be verbalized openly, and
many women were too inhibited to admit it to themselves. The study concluded that many
homemakers found an outlet for this dislike by transferring it to Saran Wrap. The frustrations
they had with the use of the product came to symbolize their frustrations with their role and
lifestyle. As a result of the study, the product was made less clingy, and nonkitchen uses were
stressed.
191
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
The underlying assumption of the clinical approach is that people often cannot or will not
verbalize their true motivations and attitudes. They may be embarrassed to reveal that they dislike cooking. Alternatively, they may have suppressed this dislike and not even be conscious of
it.25 They simply may believe that their dislike is caused by the plausible judgment that Saran
Wrap is awkward to use. The difficulty with clinical research is that true motivations are seldom
clear. Indeed, two different clinical analysts working from different theoretical backgrounds may
arrive at totally different interpretations. These problems have brought considerable disrepute to
motivation research. At present, this type of research is relegated to a distinctly secondary role;
however, projective techniques for asking indirect questions, when direct questions may not provide valid answers, are used more extensively. The following categories of projective techniques
will be discussed: (1) word association, (2) completion tests, (3) picture interpretation, (4) thirdperson techniques, (5) role playing, and (6) case studies.
Word Association. The word-association technique asks respondents to give the first word or
phrase that comes to mind after the researcher presents a word or phrase. The list of items used
as stimuli should include a random mix of such neutral items as “chair,” “sky,” and “water,” interspersed with the items of interest, such as “shopping downtown,” “vacationing in Greece,” or
“Hamburger Helper.” An interviewer reads the word to the respondents and asks them to mention the first thing that comes to mind. The list is read quickly to avoid allowing time for defense
mechanisms to come into play. Responses are analyzed by calculating (1) the frequency with which
any word is given as a response, (2) the amount of time that elapses before a response is given,
and (3) the number of respondents who do not respond at all to a test word within a reasonable
period of time.
New products tend to come from existing products; new technologies are applied to make
them faster, healthier, less expensive, and easier to use. In fact, many researchers argue that there
is no such thing as a really new product “function” because consumer needs are relatively permanent; that is, new engineering solutions or technologies periodically address the same ongoing
consumer problems. Many years ago, people cleaned their hair with water; today they do it with
some new shampoo technology. Either way, the basic function—cleaning hair—remains unchanged.
The role played by marketing research in a really new product development is limited. Consumers
simply cannot describe needs for products that do not exist. Fifty years ago, for example,
how could marketing researchers have identified the huge demand among drivers for automatic transmissions?
Traditional marketing research methods have been largely confined to asking people about
problems with current products, watching them use these products, and giving them new prototypes for extended-use tests.
Researchers Jeffrey Durgee, Gian Colarelli O’Connor, and Robert Veryzer describe a new
method for identifying new consumer or industrial product functions. They asked consumers to
rate “mini-concepts,” novel verb-object combinations that describe possible new functions for
products. (A new function for a vacuum cleaner, for example, might be “deodorize rug.”)
Whether the technology is available to perform these new functions is immaterial. The aim
of the mini-concepts method is simply to generate interesting, “really new” function ideas and
assess buyer responses.
The authors tested the mini-concept method among 30 middle-aged mothers regarding needs
and opportunities for new food-related products. Their results indicated latent needs for products that would educate them about food, enhance their enjoyment of food, and enable them to
prepare foods that have healing and energizing properties.
Using this method, marketing researchers can identify opportunities for really new functions in
any product category. Current technologies can then be applied—or new technologies developed—
to address and fulfill these needs.
Their study describes a new method for addressing this traditional limitation of marketing
research. Asking consumers to rate novel verb-object combinations generated a number of interesting really new function ideas and allowed researchers to assess buyer responses.
192
QUALITATIVE RESEARCH METHODS
Using the method, they elicited potential new functions for food-processing technology. Top
scorers included “spend less money on food,” “alert to spoiling food,” “learn more about food,”
and “calculate calories of food.”26
The result of a word-association task often is hundreds of words and ideas. To evaluate quantitatively the relative importance of each, a representative set of the target segment can be asked
to rate, on a five-point scale, how well the word fits the brand, from “fits extremely well” to “fits
not well at all.” It is also useful to conduct the same associative research on competitive brands.
When such a scaling task was performed for McDonald’s on words generated from a wordassociation task, the strongest associations were with the words Big Macs, Golden Arches,
Ronald, Chicken McNugget, Egg McMuffin, everywhere, familiar, greasy, clean, food, cheap,
kids, well-known, French fries, fast, hamburgers, and fat. In the same study, Jack-in-the-Box had
much lower associations with the words everywhere, familiar, greasy, and clean, and much
higher associations with tacos, variety, fun, and nutritious.27
The word-association technique has also been particularly useful for obtaining reactions to
potential brand names. Consumers associate a brand with (1) product attributes, (2) intangibles,
(3) customer benefits, (4) relative price, (5) use/application, (6) user/customer, (7) celebrity/
person, (8) lifestyle personality, (9) product class, (10) competitors, and (11) country/geographic
area.28 This technique is being used extensively to explore these associations. Word association
has also been used to obtain reactions to and opinions about advertising slogans. For example,
Bell Telephone found that one theme for advertising, “The System Is the Solution,” triggered
negative, “Big Brother is watching you” reactions among some people.
Completion Tests. The simplest completion test involves giving a respondent an incomplete
and ambiguous sentence, which is to be completed with a phrase. The respondent is encouraged
to respond with the first thought that comes to mind. Sentences are usually in the third person
(“The average person considers television _____ .” “People drawing unemployment compensation are _____ .”), but may refer directly to the object or activity (“Insurance of all kinds is _____ .”)
The completion test can be expanded readily to involve the completion of a story presented as an
incomplete narrative or simply as a cartoon. In one such study, people were shown a crude picture
of two women shopping in a supermarket, each pushing a shopping cart (Figure 8-1). When told
that one woman is purchasing dry soup mix, they were asked to tell a story about her and to
FIGURE 8-1
Two women shopping.
SOURCE: Courtesy Sidney J. Levy, Eller College of Management, University of Arizona.
193
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
describe what she is saying to the second woman. They were also asked to tell what the second
woman is like. Based on this, user profiles for the soup mix were developed.29
Picture Interpretation. The picture interpretation technique is based on the Thematic Apperception Test (TAT). The respondent is shown an ambiguous picture in the form of a line drawing, illustration, or photograph, and asked to describe it. This is a very flexible technique, for the
pictures can be adapted readily to many kinds of marketing problems. An example of picture
interpretation was a study that showed the respondents two scenes.30 One involved a break after
a daytime hike on a mountain; the other showed a small evening barbecue with close friends.
During the scene the beer served was either Coors or Lowenbrau.
Respondents were asked to project themselves into the scene and indicate, on a five-point
scale, the extent to which they would feel “warm,” “friendly,” “healthy,” and “wholesome.” The
study was designed to test whether the advertising of Coors and Lowenbrau had established
associations with their use-contexts—Coors with hiking, wholesomeness, and health, and
Lowenbrau with a barbecue-type setting, friends, and warmth. The results showed that Coors
was evaluated higher in the mountain setting and Lowenbrau in the barbecue setting, as
expected, but that the other (word) associations were not sensitive (related) to the setting. For
example, in the hiking context, Coors was higher on the “warm” and “friendly” dimensions, as
well as on “healthy” and “wholesome.”
Third-Person Techniques. By asking how friends, neighbors, or the average person would
think or react in the situation, the researcher can observe, to some extent, the respondents
projecting their own attitudes onto this third person, thus revealing more of their own true
feelings. Magazines use this technique to identify which articles to feature on the cover, to
stimulate newsstand sales. Direct questioning as to the articles of greatest interest to the
respondent tends to be confounded by socially desirable responses. For example, articles on
complex issues of foreign affairs are rated highly interesting to the respondent during direct
questioning, but are not thought to be of interest to the neighbors.
Another variant of this technique provides a shopping list or a description of a person’s
activities, and asks respondents to describe the person. The respondents’ attitudes toward the
activities or items on the list will be reflected in their descriptions of the person. Usually, two
lists are prepared and presented to matched sets of respondents; these could be grocery shopping lists, in which all items are identical except that Nescafe instant coffee on the first list is
replaced by Maxwell House (drip grind) coffee on the second list;31 or the contents of a billfold,
which differ only in the inclusion of a Bank Americard on one list. Differences in the descriptions
attributed to the two lists can reveal the respondents’ underlying attitudes toward the product or activity that is being studied.
Role Playing. In role playing, a respondent assumes the role or behavior of another person,
such as a salesperson in a store. This person then can be asked to try to sell a product to consumers, who raise objections. The method of coping with objections may reveal the respondents’
attitudes, if they project themselves fully into the role playing without feeling uncomfortable or
embarrassed.
Another technique with similar expressive objectives is the role rehearsal procedure used as
part of a focus-group discussion. The participants in a focus group are encouraged, by offering
them an incentive, to alter their behavior pattern in some extreme way. Abelson describes a study
in which homemakers were asked to serve chicken to their families three times a week for a year,
in return for $15.00 a week and an agreement not to tell the family about the arrangement.32 The
reaction of the participants to this offer, as they “rehearsed” the problems and objections they
would likely encounter, gave useful insights into their own attitudes toward chicken. This technique is used toward the end of the focus-group session, and when the exercise is finished,
respondents must be told that the offer was fictional.
194
QUALITATIVE RESEARCH METHODS
Case Studies. A case study, in the research sense, is a comprehensive description and analysis
of a single situation. The data for a case study usually are obtained from a series of lengthy,
unstructured interviews with a number of people involved in the situation, perhaps combined
with available secondary and internal data sources.
Case studies are very productive sources of research hypotheses. This approach was used
by a food company to suggest the attributes that might characterize successful district sales
managers. A successful and an unsuccessful manager from otherwise similar territories (that is,
the territories had similar market structure, potential, and competitive situations) were studied
closely for two weeks. They were interviewed, observed during sales calls and trips with their
salespeople, and given a series of personality tests. The differences were used to develop a series
of surveys that were administered to all the managers.
In some circumstances a case study may be the only way to understand a complex situation. For example, the decision-making processes in large organizations may be imperfectly
understood by a single participant. This problem makes it difficult to understand the sequence
of decisions leading to, for example, the choice of a telephone service that customers use to
call for reservations or information, or to place purchase orders. A telecommunications manager may be simply a technical consultant on the telephone system for the using company and
not know how the system is used in the business. The functional managers in marketing or
operations actually may make the decision to offer the service to customers but not know the
intricacies of the switching network. To get a picture of the company’s use of the service, all
parties must be interviewed. Marketing Research in Action 8-10 makes a case for case studies.
Other Projective Techniques. Many other projective techniques have been developed and used
in recent years. BBDO Worldwide has developed a trademarked technique called Photo Sort.
Consumers express their feelings about brands through a specially developed photo deck showing pictures of different types of people, from business executives to college students. Respondents connect the people with the brands they think they use. Another photo-sort technique
called the Pictured Aspirations Technique (PAT) has been created by Grey Advertising. This
device attempts to uncover how a product fits into consumers’ aspirations. Consumer drawings
MARKETING RESEARCH IN ACTION 8-10
A Case for Case Studies
Which would you rather bet your company’s strategy on: what consumers say or what they do? Surveys,
focus groups, and mall intercepts attempt to develop an understanding of customers’ motivations by collecting reactions to researchers’ questions. The major criticism of these methods is that what they gather
are mere opinion statements that don’t reveal actual behavior. Instead of gathering poorly considered
opinion statements, the case-study approach builds insight into marketing behavior by pursuing—and
verifying—the stories behind specific recent purchases in a given product category. It builds the full stories of how 50 supermarket shoppers selected their peanut butter last Tuesday, or how 25 companies
replaced their PBX systems last month.
Key arguments for the case-study approach are
■ Case studies uncover motivations through demonstrated actions, not through statements of opinions.
■ They’re conducted in the surroundings where a product is bought or used to achieve greater immediacy (and accuracy) of response.
■ They use observation and documentation to stimulate questions and corroborate responses.
■ They access multiple decision makers.
■ They require the talents of “marketing detectives” rather than “census takers.”
SOURCE: “Study What People Do, Not What They Say,” Marketing News, January 6, 1992, pp. 15, 32.
195
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
MARKETING RESEARCH IN ACTION 8-11
Focus-Group Projective Techniques
When conducting a focus group, one goal for the moderator is to get to deeper feelings, emotions, and
values. Companies are seeking the elements that drive consumers in order to create successful marketing communication programs. The following are some projective techniques that are often used to elicit
findings beyond the top-of-mind reactions of focus-group members:
■ Animal exercise: In order to discover their underlying feelings about a product, company, or person,
researchers often ask people to think of the item as an animal and have them write down the animal’s
name and why they selected it. For example, people may think of a nonprofit organization as a
Labrador than as simply a dog because of the animal’s personality traits, such as loyalty, caring, and
intelligence.
■ Mindmapping: Researchers often ask respondents to write down all the key words that come to mind
when thinking of the central idea and then have them connect the key words in complete sentences.
When done within a time limit, this technique can produce results quickly and generate discussion
topics for the focus group.
■ Projective drawing: Researchers often ask participants to draw anything they want so as to gather visual
insights about their underlying thoughts about an issue. People generally start drawing with a simple
figure and then fill in details about that figure’s life, such as name, age, facial features, clothing, and
family. Researchers then ask the participants such questions as: What is his daily life? What are the
impressions she conveys to people? What does his family think of him? This technique can be used to
elicit consumers’ perception of a certain product (such as Miller Lite or Bud Light).
■ Picture sort: Similar to the animal exercise, researchers display pictures that represent a range of emotions. These pictures could be images of animals, people, or landscapes. Participants select the picture that reminds them of the issue and then are asked why they selected that picture and what
impression they get from it.
■ Debate session: Researchers can use debates to research products or organizations, especially when
they are connected to highly emotional topics such as war or environmental issues. Participants are
divided into two groups and assigned one side of the issue. Each group is given a certain amount of
time to develop its best arguments. The moderator then has each group appoint a spokesperson but
allows any group member to chime in during the debate. In addition, the moderator asks questions,
directs the debate, and listens for each group’s key messages to determine their underlying values or
feelings about a certain issue.
SOURCE: Bill Dalbec, “Stage an Intervention for the Focus Group,” Marketing News, February 2001, p. 46.
are used to unlock motivations or express perceptions. Researchers ask consumers to draw what
they are feeling or how they perceive an object. An example of how Internet technology is used
to elicit the inner feelings, not simply the top-of-mind reactions of consumers, is given in Marketing Research in Action 8-11.
Most people make rational, conscious, purchase decisions and can accurately report the reasons for their decisions when asked. However, many decisions appear to result from preconscious
drives that can be either expressed or repressed. When a drive is repressed, a consumer may not
be able to access the reason for a particular purchase decision. A technique called the implicit model
provides a solution to this type of a problem.
In one study, a respondent, discussing the brand of interest, had only positive comments
regarding the brand. Then a set of animal photographs was presented and the respondent was
asked to select the animal that best represented this brand. A very different picture was painted
of the brand. The photo chosen was described as a sly, wily, cunning, and even sneaky animal.
In other words, this animal was a pretty, nice, neat-looking animal that was really a fairly mean,
predatory creature stealing its prey in the night.33
196
OBSERVATIONAL METHODS
Limitations of Qualitative Methods
Most of the limitations of these qualitative methods stem from the susceptibility of the results to
misuse, rather than their inherent shortcomings. There is a great temptation among many managers to accept small-sample exploratory results as sufficient for their purposes, because they are
so compelling in their reality. The dangers of accepting the unstructured output of a focus group
or a brief series of informal interviews are twofold. First, the results are not necessarily
representative of what would be found in the population, and hence cannot be projected. Second,
there is typically a great deal of ambiguity in the results. The flexibility that is the hallmark of these
methods gives the moderator or interviewer great latitude in directing the questions; similarly, an
analyst with a particular point of view may interpret the thoughts and comments selectively to
support that view. In view of these pitfalls, these methods should be used strictly for insights into
the reality of the consumer perspective and to suggest hypotheses for further research.
OBSERVATIONAL METHODS 34
Observational methods provide a researcher with information on current behavior. Too often, this
limitation becomes an excuse for not considering observational methods; because many
researchers do not use these methods, they may not appreciate their considerable benefits.
Nevertheless, there are strong arguments for considering the observation of ongoing behavior as
an integral part of the research design. Some of these are the following:
■ Casual observation is an important exploratory method. Managers continually monitor such
variables as competitive prices and advertising activity, the length of lines of customers waiting for service, and the trade journals on executives’ desks, to help to identify problems and
opportunities.
■ Systematic observation can be a useful supplement to other methods. During a personal interview, the interviewer has the opportunity to note the type, condition, and size of the residence,
the respondent’s race, and the type of neighborhood with regard to mixed types and qualities
of homes and apartments. Seldom is this data source adequately exploited in surveys.
■ Observation may be the least expensive and most accurate method of collecting purely behavioral data such as in-store traffic patterns or traffic passing a certain point on a highway system.
Thus, people’s adherence to pedestrian safety rules before and after a safety campaign can be
measured most easily by counting the number of people who cross against the light or outside
the crosswalks.
■ Sometimes observation is the only research alternative. This is the case with physiological phenomena or with young children who cannot articulate their preferences or motives. Thus, the
Fisher Price Company operates a nursery school in a residential area as a means of fieldtesting potential new toys.
Direct Observation
Direct observation is frequently used to obtain insights into research behavior and related issues,
such as packaging effectiveness. One firm uses an observer, disguised as a shopper, to watch grocery store shoppers approach a product category; to measure how long they spend in the display
area, and to see whether they have difficulty finding the product; and whether the package is
read, and if so, whether the information seemed hard to find. This kind of direct observation can
be highly structured, with a detailed recording form prepared in advance, or very unstructured.
When making an unstructured observation, the observer may be sent to mingle with customers
in the store and look for activities that suggest service problems. This is a highly subjective task,
197
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
because the observer must select a few things to note and record in varying amounts of detail.
This inevitably will draw subjective inferences from the observed behavior. For example, just what
was meant by the frown on the face of the shopper waiting at a cash register?
Regardless of how the observation is structured, it is desirable that the respondents not be
aware of the observer.35 Once conscious of being observed, people may alter their behavior, but
in very unpredictable ways. One-way mirrors, disguises, and cameras are some of the common
solutions. Care should be taken, however, that there is not an invasion of privacy. Marketing
Research in Action 8-12 explains Procter & Gamble’s experience in observational methods in
developing new products.
MARKETING RESEARCH IN ACTION 8-12
P&G Sweeping up a Market
When Procter & Gamble launched its Swiffer mop in 1999, it was introduced as a dry mop to replace
the traditional mop and bucket. However, with roughly 75 percent of American homes being covered
with carpet, P&G now hopes to push the popular Swiffer mop onto the carpets of American homes,
sidelining the traditional vacuum. Following this, studies were conducted to identify the tasks that consumers found unpleasant when it comes to cleaning their homes. For this purpose, visits were made to
consumer homes during October 2003 to identify the disagreeable, awkward, or inconvenient issues
faced by consumers about cleaning. The study included observing carefully how consumers cleaned the
floor, dusted their furniture, and vacuumed the carpet. Even dirt was scattered around the living room
rugs to see how consumers dealt with it.
The studies threw light on various aspects of cleaning for P&G. Their research showed that three
out of every four times a consumer pulled out a vacuum cleaner, it was to tackle a little debris on a
small portion of the carpet—things like leaves, grass, and nail clippings that do not require the full
force of a traditional vacuum. Further, they noticed that a certain portion of the population was allotting less time to one big weekly cleaning session and instead were cleaning in a piecemeal fashion during the week, in between other errands or tasks. They identified several “compensating behaviors” that
consumers adopted in order to tackle the inconvenience of vacuuming. Some left their vacuum cleaner
in a corner of the room, always plugged in, to avoid the hassle of putting the machine away and getting it out again. Others bought several vacuums, one for each level of the house, to avoid carrying the
machine up and down the stairs. Yet another group simply avoided the task altogether by buying carpets that did not show much dirt. P&G concluded that “the setup and breakdown time” of the traditional vacuum cleaner was one of the things that kept people from cleaning.
Based on the results of these studies, P&G planned to launch a cordless carpet sweeper in August
2005. The Swiffer CarpetFlick, their seventh major Swiffer brand extension, is designed to clean up
small patches of dirt, grass, and debris spilled on carpets, and keep consumers from dragging out
their traditional vacuum cleaners for the small jobs. The new CarpetFlick resembles an old-fashioned
manual sweeper, but operates on a different principle. The plastic on the base of the Swiffer
CarpetFlick pushes down on crumbs or dirt, and then flicks the particles up through an opening in
the sweeper into a cavity that contains a sticky adhesive sheet, where the debris is captured. The
adhesive is double-sided, and the bottom side helps pick up fuzz and lint that is not “flicked” into
the body. Like other Swiffer products, the adhesive is disposable, and replaceable. The new Swiffer
has a tinted orange window through which consumers can see the debris they are collecting. The
tinting keeps the refuse from being too visible, but allows people to keep track of how much dirt is
stuck to the adhesive sheet. This way, consumers need not have to deal with the “bagful of dirt.” At
the same time, as they can see the dirt they are collecting, it provides consumers with a signal that
they are getting the floor cleaned.
Procter & Gamble hopes the new device can extend the run of the Swiffer brand, which has benefited from consumers who are lazy in their cleaning habits. The “Quick-clean” products have become
the fastest-growing category of the $3 billion-plus cleaning-products market. Procter & Gamble estimates
the niche, which includes products like the Swiffer, Clorox Co.’s Toilet Wand, and S.C. Johnson & Son
Inc.’s Pledge Grab-it, at $800 million, and is growing at about 7 percent a year. Since, most of the products in this category have replacement cleaning elements that can be thrown away once they are dirty,
consumers now need not bother about dealing with dirty appliances. They’re also appealing to manufacturers, who can charge a premium for the refills.
198
SOURCE: Adapted with permission from Sarah Ellison, “Studying Messy Habits to Sweep up a Market,” The Wall Street Journal, July 14, 2005.
OBSERVATIONAL METHODS
Contrived Observation
Contrived observation can be thought of as behavioral projective tests; that is, the response of
people placed in a contrived observation situation will reveal some aspects of their underlying
beliefs, attitudes, and motives. Many direct-mail offers of new products or various kinds of books
fall into this category, as do tests of variations in shelf space, product flavors, and display locations. The ethics of such offers can be very dubious, as in the example where a manufacturer
decides to produce a product only after receiving an acceptable number of orders from a directmail advertisement.
A variant of this method uses buying teams, disguised as customers, to find out what happens during the normal interaction between the customer and the retailer, bank, service department, or complaint department. This method has provided useful insights into the discriminatory
treatment of minorities by retailers, and the quality of public performance by employees of government agencies, banks, and airlines. One is hard pressed to think of other ways of finding out
about the knowledgeability, helpfulness in meeting customers’ needs, and efficiency of the staff.
Clouding this picture are some serious, unresolved questions of ethics.
Content Analysis
Content analysis is an observation technique used to analyze written material into meaningful
units, using carefully applied rules.36 It is defined as the objective, systematic, and quantitative
description of the manifest content of communication. It includes observation as well as analysis. The unit of analysis may be words, characters, themes, space and time measures, or topics.
Analytical categories for classifying the units are developed, and the communication is broken
down according to prescribed rules. Marketing research applications involve observing and analyzing the content or message of advertisements, newspaper articles, television and radio programs, and the like. For example, a study hypothesized that because of the growing number of
elderly Americans, advertisers would use more elderly models in their promotions. After a content analysis of all the advertisements, the researchers found that the use of elderly people in
advertisements has indeed increased.
Physical Trace Measures
Physical trace measures involve recording the natural “residue” of behavior. These measures are
rarely used, because they require a good deal of ingenuity and usually yield a very gross measure. When they work, however, they can be very useful.37 For instance, (1) the consumption of
alcohol in a town without liquor stores has been estimated from the number of empty bottles in
the garbage;38 (2) an automobile dealer selected radio stations to carry his advertising by observing the most popular dial settings on the radios of cars brought in for servicing; (3) one magazine
readership research method employs small glue spots in the gutter of each page spread of a magazine, so broken glue spots are used as evidence of exposure; and (4) a museum gauges the popularity of individual exhibits by measuring the rate of wear on the floor tiles in front of the exhibit
and by the number of nose smudges on the glass of the case around the exhibit.
The home audit approach to purchase panels (described in Chapter 6) is yet another type
of physical trace measure. The auditor describes the inventory in several prespecified categories.
This method is not very useful if used on a one-shot basis, for it then requires a very tenuous
assumption that possession indicates purchase and usage. However, if the inventory is made over
an extended period and supplemented with a record of cartons and wrappers, an indication of
the rate of purchase is possible.
Empathic Interviewing39
Empathic interviewing is an exploratory form of research that draws from the wisdom of
sociology, psychology, market research, and anthropology to help researchers probe beneath
199
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
generalizations and identify the social factors that influence consumer behavior. In understanding how complex decisions and behaviors emerge, empathic interviewing can supplement or replace traditional research methods. Some guidelines in conducting empathic
interviews are as follows:
1.
2.
3.
4.
Imagine yourself in the person’s situation.
Avoid self-referencing.
Gently challenge generalizations by asking for specific examples.
Ask open-ended, nonleading questions that start with how, what, and why.
Humanistic Inquiry
Humanistic inquiry is a controversial research method that relies heavily on observation, but
is now being used in marketing with increasing frequency.40 The humanistic approach advocates immersing the researcher in the system under study rather than as in the traditional
scientific method, in which the researcher is a dispassionate observer. Throughout the immersion process, the humanistic researcher maintains two diaries, or logs. One is a theory construction diary that records in detail the thoughts, premises, hypotheses, and revisions in the
researcher’s thinking. The second set of notes the researcher maintains is a methodological
log. Detailed and time-sequenced notes are kept on the investigative techniques used during
the inquiry, with special attention to biases and distortions a given technique may have introduced. To access whether the interpretation is drawn in a logical and unprejudiced manner
from the data gathered and the rationale employed, humanistic inquiry relies on the judgment of an outside auditor or auditors. Marketing Research in Action 8-13 illustrates humanistic inquiry research.
Behavior-Recording Devices
Various behavior-recording devices have been developed to overcome particular deficiencies in
human observers. The most obvious example is the traffic counter, which operates continuously
MARKETING RESEARCH IN ACTION 8-13
Consumption Values and Lifestyles of WASPs
The objective of the study was to interpret the consumption values and lifestyles of WASP (White, AngloSaxon, Protestant) consumers. Since this group of consumers is very large, the researcher decided to narrow down his scope geographically to communities along the eastern seaboard. The researcher made
field trips to such locales as Richmond, Virginia; Charleston, North Carolina; Connecticut; and Maine.
During these trips observations were made in residential neighborhoods, shopping areas, and other
places. Whenever the researcher had personal contact in the community, he lived in the local household
observing and participating in their daily rounds; attending church services and community meetings;
and engaging in shopping activities and meeting and talking with other residents. After every meeting
or contact episode, the researcher made a note of the thinking, transformation, and conclusion he made
regarding WASPs’ consumption values and lifestyle in his theory construction diary, while he made a
note of the happenings, the time and date when a certain event occurred in his methodological log. After
18 months of participating in the WASP lifestyle and meticulous record keeping, a pattern began to
emerge. The pattern was formed around the central values that WASP consumers had expressed verbally to the researcher in several contexts and other documentary evidence. The researcher was suddenly able to grasp that the same core values (practicality, conservatism, individual responsibility, self
control, etc.) were expressed in all aspects of the consumers’ lifestyle—from clothing and automobile
preference to leisure activities.
200
SOURCE: Elizabeth C. Hirschman, “Humanistic Inquiry in Marketing Research: Philosophy Method and Criteria,” Journal of Marketing Research, August 1986,
pp. 237–249.
OBSERVATIONAL METHODS
without getting tired, and consequently is cheaper and probably more accurate than humans.
For the same reasons, as well as for unobtrusiveness, cameras may be used in place of human
observers. Someone still has to interpret what is recorded on the film, but the options exist of
sampling segments of the film, slowing the speed, or having another observer view it for an
independent judgment.
Of the mechanical devices that do not require respondents’ direct participation, the A.C.
Nielsen “people meter” is best known. The people meter is attached to a television set to record
continually to which channel the set is tuned. It also records who is watching. Arbitron recently
developed a pocket people meter that is no larger than an electronic pager and can recognize
the unique code that broadcasters embed in the soundtrack of radio or television programs.41
The latest trend is to develop “passive people meters.” These meters will record an individual’s
broadcast viewing and listening without his or her intervention. Mediacheck has introduced
one such system. It has two units: a set-top meter that detects audio codes to identify programming being watched or listened to, and a device such as a pendant or a wristwatch worn
by the person being measured. The main drawback of this system is that it requires that all programming carry a nonintrusive audio code that can be recorded by the meter system. Nielsen
is also developing its own passive people meter. It will recognize a person entering the room
and also keep track of the person’s activities while he or she watches TV. Technological advances
such as the Universal Product Code (UPC) have had a major impact on mechanical observation. The UPC system, together with optical scanners, allow for mechanical information collection regarding consumer purchases, by product category, brand, store type, price paid, and
quantity.
Some types of observation are beyond human capabilities. All physiological reactions fall
into this category. Therefore, devices are available to measure changes in the rate of perspiration
as a guide to emotional response to stimuli (the psychogalvanometer), and changes in the size
of the pupils of subjects’ eyes, which are presumed to indicate the degree of interest in the stimulus being viewed (the pupilometer). These devices can be used only in laboratory environments,
and often yield ambiguous results.
Experience with eye-movement recorders has been more successful. This device records the
experience of viewing pictures of advertisements, packages, signs, or shelf displays, at a rate of
30 readings per second. The recorded eye movements show when the subjects starts to view a
picture, the order in which the elements of the image were examined and reexamined, and the
amount of viewing time given each element. One application is for testing the visual impact of
alternative package designs. An example of how technology can help in data collection using
eye-movement recorders is provided in Marketing Research in Action 8-14.
MARKETING RESEARCH IN ACTION 8-14
Tracking Shoppers Virtually
Kimberly-Clark Corp. has developed a new tool that tracks retina movements and records every glance
of the consumer. Using the tool virtual shopping aisles was created wherein consumers are requested to
shop for products. The shopping behavior was then captured by the devices and the observations are
used for understanding consumer behavior.
Kimberly-Clark hopes these virtual shopping aisles will help it better understand consumer behavior and make the testing of new products faster, more convenient and more precise. The mobile testing
unit is usually based in a new high-tech studio that Kimberly-Clark completed in May 2007 in
Appleton, Wisconsin. The room also features a U-shaped floor-to-ceiling screen that re-creates in vivid
detail interiors of the big retailers that sell the company's products—a tool that the company will use in
presentations to executives in bids to win shelf space. A separate area is reserved for real replicas of
store interiors, which can be customized to match the flooring, light fixtures and shelves of retailers such
as Target Corp. and Wal-Mart Stores Inc. Kimberly-Clark uses the data from its virtual-reality tests with
consumers to tout how products in development perform.
SOURCE: Ellen Byron, “A Virtual View of the Store Aisle,” Wall Street Journal, October 3, 2007.
201
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
MARKETING RESEARCH IN ACTION 8-15
New Technology Collects Media Behavior Data
Systems for measuring radio and TV audiences haven’t improved much over the past few decades. Audiences are measured by small, representative panels. TV panels use set-top boxes. Radio panelists are asked
to record listening habits in a diary. Researchers use highly refined methods to create a small sample of individuals whose behaviors and attitudes represent everyone. The problem is understanding the media habits
of the panel members: to accurately determine what they watch and listen to, when, and for how long.
Germany’s largest market research company, GFK, presented a new technology to measure what TV
and radio shows a person watched or listened to by using a wristwatch named Radiocontrol. Several
thousand panel members are given watches to wear for two weeks. The watch takes a four-second “fingerprint” of ambient sound every minute. At the end of two weeks, the watch is returned and the sound
fingerprint is matched against a database of radio and TV programming to determine exactly what a person watched or listened to. The sample is then projected out to a national audience.
The watch is a highly sophisticated instrument, containing a motion detector and a thermostat to
determine whether it is worn at any given moment. It’s sleek and stylish but automatically stops telling
time after two weeks, which prompts the wearer to return it to the research company.
The ability to accurately track people’s real-world behaviors is of tremendous interest to marketers.
When this type of technology is married with attitudes and purchase behavior, marketers will have a
direct link between what people think and buy and the advertising they’re exposed to. It promises to
create enormous advertising efficiencies.
SOURCE: Jeffrey Graham, “Tracking Moves Offline,” http://www.clickz.com/clickz/column/1711679/tracking-moves-offline, November 27, 2002.
Voice-pitch analysis examines changes in the relative vibration frequency of the human voice to
measure emotion.42 In voice analysis, the normal or baseline pitch of an individual’s speaking voice
is charted by engaging the subject in an unemotional conversation. The greater the deviation from
the baseline, the greater is said to be the emotional intensity of the person’s reaction to a stimulus.
For example, voice-pitch analysis has been used in package research, to predict consumer brand
preference for dog food.43 An example of how technology can help in data collection through direct
link is provided in Marketing Research in Action 8-15.
Limitations of Observational Methods
The vast majority of research studies use some form of questionnaire. Observation methods, despite
their many advantages, have one crucial limitation: They cannot observe motives, attitudes, or
intentions, which sharply reduces their diagnostic usefulness. To be sure, these cognitive factors
are manifested in the observed behavior, but so are many other confounding factors. For example,
the Zippo Lighter Company seemingly has a valuable measure of advertising effectiveness in the
volume of its lighters sent in for repair. Despite the mention of the free repair privilege in the advertising, it is questionable whether such a measure can unambiguously test for impact.
Observational methods suffer other limitations as well. They are often more costly and timeconsuming, and may yield biased results if there are sampling problems or if significant observer
subjectivity is involved. However, these biases usually are very different in character from those
that affect obtrusive questionnaire methods. This is one of the underexploited strengths of observation methods: they help to increase our confidence in questionnaire measures if they yield similar results when used as a supplement.
RECENT APPLICATIONS OF QUALITATIVE
AND OBSERVATIONAL METHODS
202
In recent years many companies, across various industry verticals, have applied qualitative and
observational techniques to acquire vital information about marketing problems. Here are a few
RECENT APPLICATIONS
OF
QUALITATIVE
AND
OBSERVATIONAL METHODS
MARKETING RESEARCH IN ACTION 8-16
Bertolli—Now Serving
In 2007, Unilever’s Bertolli brand that includes pasta sauces and olive oils, zeroed in on a market segment that did not have their presence—the prepared meals segment consisting of frozen meals. Market
research reports identified that with the growth of average work hours by 26 percent in 2006, the frozen
meals that Americans eat have increased by 47 percent. A majority (75%) of this $5.7 billion market comprised of single-serve options and about 20 percent of the market was made up of multi-serve frozen
entrées and dinner. It was this 20 percent that Bertolli was interested in as this market was growing at
an impressive 16 percent between 2003 and 2005.
Bertolli had some questions that helped them identify their brand strategy—Is it a big space or a growing space, or both? Do we have a brand that can play and compete in that space? And, if so, what difference can we make? Based on the findings, Bertolli decided to compete in complete meal solutions market and the result was the launch of “Dinner for Two” line of products.
Before launching this product line, Bertolli turned to ethnographic research in 2000 to help design
meals that would stand out from competitors—that would mark its products as an upscale, luxurious
end to a busy day, not just a convenience. They conducted observational studies which had researchers
accompanying participating families to grocery stores to observe their shopping habits and, watching
them cook and eat frozen meals. This helped them identify areas of complaint and dissatisfaction. They
learnt that polybag meals already on the market had mushy pasta and did not taste fresh. Participants
also complained that the packaging was unappealing and seemed downscale. Meanwhile, other
researchers visited traditional Italian restaurants in and popular Italian chain restaurants to see what
kinds of ingredients chefs used in their dishes.
With the market research complete, Bertolli then concentrated on the packaging of the meals. Designers placed a photograph of the cooked entrée against a sepia-toned background of the hills of Tuscany.
There’s also a glass of wine on the package, meant to underscore the idea that the food inside is on par
with a restaurant meal.
The efforts paid off for Unilever. Since launching nationwide in January 2005, Bertolli-brand frozen
entrees have garnered a 38.6 percent share of the multi-serve polybag market, outperforming both
Stouffer’s and Birds Eye. In U.S. grocery stores, Bertolli accounts for eight of the top 11 best-selling polybag dinners, according to AC Nielsen. Meanwhile, the line has become a new growth driver for Unilever
with sales rising more than 185 percent, from $70 million in its first year to a projected $200 million plus
in 2007.
SOURCE: Jeff Borden, “Bertolli’s Big Bite,” Marketing News, October 1, 2007.
examples of companies who have adopted these techniques. Marketing Research in Action 8-16
describes a brand that used observational techniques to identify a new market.
■ Maritz Marketing Research, Inc. (St. Louis), says its Virtual Customers system for evaluating service quality goes beyond typical mystery shopping programs because of its integrated
approach, customer certification process, and high-tech project management system. Virtual
customers (for example, mystery shoppers) receive job-specific training and periodic
updates during the certification process, said Al Goldsmith, director of virtual customer
research at Maritz. The emphasis on training is needed because “without it, each person
would be unlikely to role-play adequately enough to obtain reliable data.” Maritz uses a
high-tech project management system to track the skill level and demographic information
on each virtual customer. A database shows each person’s shopping performance and other
pertinent information to ensure that the right people are assigned to each project. The same
system can produce real-time feedback on the status of the project, the number of locations
visited, and current ratings compared with ratings of previous visits.
■ On-site observation is an underutilized method that has tremendous potential for helping marketers understand consumers’ thinking. It is as valuable as traditional qualitative techniques
203
CHAPTER 8 INFORMATION COLLECTION: QUALITATIVE
AND
OBSERVATIONAL METHODS
such as focus groups and in-depth interviews. On-site observation gives researchers the chance
to observe and learn about consumer purchasing decisions as they are being made. Observing consumer decisions as they choose products from a shelf or casually asking them to
explain their preferences can lead to a wealth of qualitative information. The best approach
in conducting on-site observation is by placing researchers in a supermarket and letting them
present themselves as shoppers who need advice from other shoppers in making a purchasing decision. The shoppers in question will give spontaneous responses on the product that
are noteworthy for marketing analysis.
A skilled researcher can do about 35 interviews a day in several supermarkets with minimal overhead costs and an excellent yield for qualitative work. It requires individuals who
are outgoing and are able to listen—not just hear what people say. A person with these traits
can then be taught how to obtain relevant answers, absorb information, evaluate responses,
and avoid asking leading questions. Field observations also require tremendous patience, discipline, and enthusiasm.44
Summary
Exploratory research is an essential step in the development of a successful research study. In essence, this
kind of research is insurance that major elements of the problem or important competing hypotheses will
not be overlooked. It further ensures that both the manager and the researcher will see the market through
the consumer’s eyes. Fortunately, research design is an iterative and not a sequential process, so major initial oversights are not necessarily irreversible. In particular, the exploratory technique of semistructured interviews should be reemployed later when the structured study is pretested. Properly handled, a pretest should
provide opportunities for respondents to express their frustrations with the specific questions, as well as
identify deficiencies in the scope of the questions.
In this chapter we also discussed observational methods. These are useful during the exploratory
stage of the research design but are even more valuable as a data collection method. The advantages of
observational methods will become even more apparent in Chapter 9, as we examine the errors that are
inherent when an interviewer starts to interact with a respondent.
Questions and Problems
1. What are the significant differences between nondirective and semistructured individual interviews?
In what circumstances would a nondirective
interview be more useful than a semistructured
interview?
2. You have conducted two group meetings on the
subject of telephone answering devices. In each
group there were seven prospective users of such
devices, and in the two groups there were four
users of telephone answering services. (These
services use an operator to intercept calls and
record messages.) When the client’s new-product
development manager heard the tapes and read
the transcripts of the two meetings, the first reaction was, “I knew all along that the features I
wanted to add to our existing model would be
winners, and these people are just as enthusiastic.
Let’s not waste any more research effort on the
question of which features are wanted.” What do
you say?
204
3. There have been a number of complaints in your
city that minorities are discriminated against by
local major appliance retailers (with respect to
prices, trade-ins, sales assistance, and credit). How
would you use the techniques described in this
chapter to study this question?
4. A local consumer organization is interested in the
differences in food prices among major stores in
the area. How should it proceed in order to obtain
meaningful comparisons?
5. Toothpaste manufacturers have found consistently
that if they ask for detailed information on the frequency with which people brush their teeth, and
then make minimal assumptions as to the quantity
of toothpaste used on each occasion, as well as
spillage and failure to squeeze the tube empty, the
result is a serious overstatement of toothpaste
consumption. How would you explain this phenomenon? Would it be possible to design a study to
overcome these problems and obtain more accurate
estimates of consumption? Describe how such a
study could be conducted.
6. The school board of St. Patrick’s High School is
concerned about the rise of student violence
involving guns during the past two months. During an emergency meeting, Father Hennessy, the
school’s principal, insisted that steps be taken to
return student behavior to its traditionally exemplary level. He suggested that this rise in the level
of violence was for reasons beyond the experience
of most of the school board members, and recommended that they needed to gather some background information on the situation in order to
address the problem knowledgeably. The school
board agreed with his proposal and decided to
commission a market research study to question
students on whether they carried guns and their
underlying motivation for doing so. The research
firm, Church and Associates, has decided to use
focus groups or individual in-depth interviews to
obtain the required information.
a. Which of the above-mentioned techniques is
suitable for this purpose? Give reasons.
b. Church and Associates has recommended separate focus-group discussions involving the
faculty. How can the moderator effectively
encourage all the participants to discuss their
feelings on the topic?
7.
8.
9.
10.
c. What possible questions could be asked in an
in-depth interview?
What difficulties might be encountered when conducting a qualitative interview in an international
context?
When would you recommend observational methods? Why?
How can a pretest add to the quality of a marketing research study?
For each of the following scenarios, indicate
whether qualitative or quantitative research is
more appropriate. Also recommend a specific technique for each and justify your answer.
a. A manufacturer of herbal teas wants to know
how often and for what purposes consumers
use herbal tea products.
b. A political campaign manager wants to identify
ways of sharply increasing the amount of contributions from the population.
c. An amusement park is nearing the final stages
of designing two new rides. It wants to measure
the reactions of the public and their potential
enthusiasm for the new rides.
d. A company that manufactures potato chips has
two alternative package designs for the product
and wants to determine which design will yield
higher sales.
End Notes
1. Fred N. Kerlinger, Foundations of Behavioral Research,
3rd ed., New York: Holt, Rinehart and Winston, 1986.
2. http://www.energytrust.org, October 22, 2002.
3. David Kiley, “At Long Last, Detroit Gives Consumers
the Right of Way,” Adweek, June 6, 1988, pp. 26–27.
4. Johny K. Johansson and Ikujiro Nonaka, “Market
Research the Japanese Way,” Harvard Business
Review, May–June 1987, pp. 16–22.
5. Jeffrey F. Durgee, “Depth-Interview Techniques for
Creative Advertising,” Journal of Advertising Research,
January 1986, pp. 29–37.
6. Michael Quinn Patton, Qualitative Evaluation and
Research Methods, Newbury Park, CA: Sage, 1990.
For additional discussion of cross-cultural research
and evaluation, see Michael Quinn Patton, Culture
and Evaluation: New Directions for Program Evaluation,
San Francisco: Jossey-Bass, 1985; Walter J. Lonner
and John W. Berry, Field Methods in Cross Cultural
Research, Newbury Park, CA: Sage, 1986.
7. Bonnie Eisenfeld, “Phone interviews may garner
more data,” Marketing News, March 3, 2003; Athol
Foden, “Customer-Focused Efforts Require Listening,” Marketing News, March 3, 2003.
8. Judith Langer, “Telephone Time: In-Depth Interviews via Phone Gain Greater Acceptance,” Marketing Research Forum, Summer 1996, p. 6.
9. Michael Garee and Thomas R. Schori, “Position
Focus Group Observers for Best View,” Marketing
News, September 14, 1998, p. 29.
10. Martin Lautman, “Focus Groups: Theory and
Method,” Advances in Consumer Research 9, October
1981, p. 54.
11. See Thomas Greenbaum, “Focus Groups vs. Oneon-Ones: The Controversy Continues,” Marketing
News, September 2, 1991, p. 16.
12. Peter Tuckel, Elaine Leppo, and Barbara Kaplan,
“Focus Groups under Scrutiny: Why People Go
and How It Affects Their Attitudes towards Participation,” Marketing Research, 12, June 1992.
13. For a comprehensive listing of the providers of
focus group facilities, see “1992 Marketing News
Directory of Focus Group Facilities and Moderators,” Marketing News, January 6, 1992.
14. David W. Stewart and Prem N. Shamdasani, Focus
Groups: Theory and Practice, Newbury Park, CA:
Sage, 1990.
15. Nelson E. James and Nancy Frontczak, “How
Acquaintanceship and Analyst Can Influence
Focus Group Results,” Journal of Advertising, 17,
1988, pp. 41–48.
16. Wendy Hayward and John Rose, “We’ll Meet
Again . . . Repeat Attendance at Group Discussions—
205
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
Does it Matter?” Journal of the Advertising Research
Society, July 1990, pp. 377–407.
For more discussion, see Edward F. Fern, “The Use
of Focus Groups for Idea Generation: The Effects
of Group Size, Acquaintanceship, and Moderator
on Response Quantity and Quality,” Journal of
Marketing Research, 19, February 1982, pp. 1–13.
Naomi R. Henderson, “Trained Moderators Boost
the Value of Qualitative Research,” Marketing
Research, June 1992, p. 20.
Rachel X. Weissman, “Online or Off Target,”
American Demographics, November 1998.
Dan Prince, “Handling the Unexpected during
Focus Group,” Marketing News, September 23,
1996, p. 29.
Michael Silverstein, “Two Way Focus Groups Can
Provide Startling Information,” Marketing News,
January 4, 1988, p. 31.
“Network to Broadcast Live Focus Groups,” Marketing News, September 3, 1990, pp. 10, 47.
Cindy Miller, “Anybody Ever Hear of Global
Focus Groups?” Marketing News, May 27, 1991,
p. 14.
Harold H. Kassarjian, “Projective Methods,” in Robert
Ferber, ed., Handbook of Marketing Research, New York:
McGraw-Hill, 1974, pp. 3–87; Sidney J. Levy,
“Dreams, Fairy Tales, Animals, and Cars,” Psychology
and Marketing, 2, Summer 1985, pp. 67–82.
Robert K. Schnee, “Quality Research: Going Beyond
the Obvious,” Journal of Advertising Research, 28,
February–March 1988.
Ben Ordini, “A Novel Idea for Market Researchers,”
Insights from MSI, Marketing Science Institute,
Cambridge, MA, Fall 1997.
David A. Aaker, Managing Brand Equity: Capitalizing on the Value of a Brand Name, New York:
Macmillan, 1991.
David A. Aaker, Managing Brand Equity: Capitalizing on the Value of a Brand Name, New York:
Macmillan, 1991.
Sidney J. Levy, “Dreams, Fairy Tales, Animals, and
Cars,” Psychology and Marketing, 2, Summer 1985,
pp. 67–81.
David A. Aaker and Douglas M. Stayman, “Implementing the Concept of Transformational Advertising,” Psychology and Marketing, May–June 1992,
pp. 237–253.
This was the design of the classic study by Mason
Haire, “Projective Techniques in Marketing Research,” Journal of Marketing, 14, 1950, pp. 649–656.
However, recent validation studies have found that
differences in the two descriptions are also influenced by the relationship of the two test products
to the items in the shopping list, so the interpretation is anything but straightforward. See also James
C. Andersen, “The Validity of Haire’s Shopping List
206
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
Projective Technique,” Journal of Marketing, 15,
November 1978, pp. 644–649.
Herbert Abelson, “A Role Rehearsal Technique for
Exploratory Research,” Public Opinion Quarterly,
30, 1966, pp. 302–305.
Carol B. Raffel, “Vague Notions,” Marketing
Research, Summer 1996, pp. 21–23.
The analysis of secondary records, as discussed in
the previous chapter, is an observational method.
In this chapter, however, we are restricting ourselves to the observation of ongoing behavior.
Cliff Scott, David M. Klien, and Jennings Bryant,
“Consumer Response to Humor in Advertising: A
Series of Field Studies Using Behavioral Observation,” Journal of Consumer Research, 16, March 1990,
pp. 498–501.
An excellent summary article of content analysis is
Richard Kolbe and Melissa Burnett, “Content
Analysis Research: An Examination of Applications with Directives for Improving Research Reliability and Objectivity,” Journal of Consumer
Research, September 1991, pp. 243–250. Other
examples of content analysis are Mary Zimmer
and Linda Golden, “Impressions of Retail Stores: A
Content Analysis of Consumer Images,” Journal of
Retailing, Fall 1988, pp. 265–293; and Terence Shimp,
Joel Urbany, and Sakeh Camlin, “The Use of Framing and Characterization for Magazine Advertising
of Mass Marketed Products,” Journal of Advertising,
January 1988, pp. 23–30.
For a fuller discussion, see Sechrest Lee, New
Directions for Methodology of Behavior Science: Unobtrusive Measurement Today, San Francisco: JosseyBass, 1979.
Joseph A. Cote, James McCullough, and Michael
Reilly, “Effects of Unexpected Situations on BehaviorIntention Differences: A Garbology Analysis,”
Journal of Consumer Research, 12, September 1985,
pp. 188–194.
Peggy Lawless, Marketing News, January 4, 1999.
Elizabeth Hirschman, “Humanistic Inquiry in Marketing Research: Philosophy, Method and Criteria,” Journal of Marketing Research, August 1986,
pp. 237–249. For more detailed information, see
Yvonna Lincoln and Edward Guba, Naturalistic
Inquiry, Beverly Hills, CA: Sage, 1985.
“Arbitron to Develop ‘Pocket People Meter,’”
Marketing News, January 4, 1993.
Nancy Nighswonger and Claude Martin, Jr., “On
Voice Analysis in Marketing Research,” Journal of
Marketing Research, August 1981, pp. 350–355.
Glen Brickman, “Uses of Voice Pitch Analysis,”
Journal of Advertising Research, April 1980,
pp. 69–73.
William B. Helmreich, “Louder Than Words,”
Marketing News, March 1, 1999, p. 13.
Appendix
Myths of Qualitative Research: It’s Conventional, but Is It Wisdom?
Myth
Fact
Focus groups should never have more than six
or seven respondents, so each one gets more
“air time.”
The ideal number of respondents depends on the
type of respondents, the subject, and the
moderator’s style.
Qualitative researchers should approach each
study with a clean-slate mentality. Past
research by other companies, information about
the clients’ hunches, or internal political battles
concerning the current study will only bias the
researcher.
Background on a project’s real objectives and past
learning is an important part of the moderator
briefing. The information could certainly bias the
moderator but a good researcher knows what to
look for. The moderator can probe responses that
do not fit hypotheses, touching more briefly on
what confirms well-accepted knowledge.
Many good ideas die in focus groups because
consumers are not ready to accept new
concepts.
Qualitative research offers researchers the
opportunity to probe what underlies consumers’
reactions. If an idea meets immediate respondent
rejection, a good researcher explores the reasons
why, whether the product might fit a real need, and
what, if anything, could change the respondent’s
mind.
Focus groups should be demographically
homogeneous.
Most research budgets cannot afford the luxury of
focus groups that differ demo-graphically, with one
group per type in each market. Some diversity may
actually prove to be beneficial. The rule of thumb is
that the respondents must feel comfortable with one
another talking about the subject.
It is a mistake to ask respondents “why” they
do or feel something.
Usually the problem is not in asking the questions,
but in how the questions are formed. The questions
should not be intimidating and must show genuine
interest and a desire to understand.
Projective methods are excellent tools but they
should not take the place of direct questions. In
many areas, straightforward, nonjudgmental
questions about respondents’ feelings successfully
uncover respondents’ emotions.
Qualitative research should rely mainly on
projective techniques (such as collages,
drawings, photographs, and role-playing). These
bring out thoughts and feelings that consumers
either are not in touch with or are
uncomfortable acknowledging publicly.
Focus groups at 8 P.M. are not a good idea
because everyone is tired.
The sign of a good moderator is that he or she
says very little in a focus group.
Recruiting is fairly easy when the client
provides lists of customers to be interviewed for
focus groups and in-depth interviews.
Focus groups at 6 P.M. and 8 P.M. are necessary
because most people work during the day. The
energy level of a focus group is more indicative of
the chemistry of respondents and their interest in
the topic than the time.
The moderator’s role should be active, responsive,
and creative. That role includes guiding the group
dynamics. Moderators should offer nonleading
feedback by reflecting back to respondents what
the moderator picks up and probing further.
Recruiting from a list is not necessarily easy. Lists
typically take longer to produce than expected,
which doesn’t allow sufficient time for recruiting.
Many lists lack telephone numbers, requiring additional time and expense for phone lookups.
207
(continued)
Sending notes to the moderator during the
focus group enables clients to be involved in
the process, ensuring that the research
provides answers to their questions.
Notes brought into the interviewing room can be
quite disrupting. Notes can be a reminder that the
respondents are being observed and hence will
make them self-conscious. Also, repeated notes
raise questions in respondents’ minds about
whether the observers trust the moderator,
creating a tense and awkward situation.
Focus groups should be used rarely since
there is usually a strong personality who
dominates the discussion.
In many groups, respondents all contribute to the
discussion, although some people are more
talkative or articulate than others. Instead, other
respondents may distance themselves from that
person, feeling free to disagree. A strong
personality may speak up confidently yet not
dominate. Experienced moderators have a variety
of techniques for dealing with respondents.
The topic guide and materials shown in a
series of focus groups should be kept the same
throughout for consistency.
Qualitative research should be dynamic; new ideas
may come up, which should alter the topic guide
during the study. Changes usually mean that learning has taken place and the topic should not
constantly remain the same. The purpose of the
qualitative study is to help the client move toward
ideas that work, not to “quantify” that an idea
doesn’t work.
Qualitative research studies should always be
followed by quantitative research to test the
hypotheses.
At times, due to budget constraints, qualitative
studies are sufficient for marketing research. In
terms of quality of research, if the research of a
sizeable qualitative study is consistent and has a
“commonsense validity,” then it is unnecessary to do
quantitative research.
A discussion guide for focus groups is referred
to by some clients as a “script,” a term that
implies that the moderator should follow it
pretty much verbatim.
A topic guide should be just that—a general guide
to the topics that should be covered and the order
in which they should be covered, and some ideas
on possible probes or techniques the moderator
might use. The sequence is merely a guideline
when there is no significant research reason for
holding off on a subject until another one has been
discussed.
Qualitative research must be done in person
and through in-depth interviews to develop
rapport between interviewer and respondent
Rapport can be just as strong or stronger in
telephone interviews. Not being able to see the
interviewer can be an advantage. To build rapport
on the phone, we ease into the interview with
simple usage or attitude questions rather than the
icebreakers often asked in focus groups.
SOURCE: Judith Langer, Marketing News, March 1, 1999, Vol. 33, No. 5, p. 13.
208
Quant-Qual Combo in Understanding Customers
Recent thinking among subject-matter specialists is focused toward integrating qualitative and quantitative
methodologies in an attempt to understand customer behavioral patterns. For quite some time now, it has
been realized that neither qualitative nor quantitative research is sufficient in perfectly describing behavioral
patterns. Quantitative methods, in their effort to provide more “objective” results, tend to overlook a multitude
of factors that may actually have been influential in decision making. Most such research, based on
quantitative analysis, is done using oversimplified linear models, which make a lot of unwarranted and unrealistic assumptions. Such assumptions tend to distort facts because dimensions like self-identity could have
played major roles in channeling customer behavior. In such cases, qualitative analysis tools like ethnographies
(an anthropological tool that studies human behavior in perspective of cultures), case studies, and histobiographies (biographies in historical context) would provide a better understanding. Using both quantitative and
qualitative analyses in tandem would be the best way to understand customer needs and how to appropriately
market to these needs.
SOURCE: Demitry Estrin, “Quant-Qual Combo Aids Audience Understanding,” Marketing News, March 3, 2003, p. 56.
CASE 8-1
Mountain Bell
Telephone Company
Jim Martin, marketing research manager for Mountain
Bell, studied the final research design for the hospital
administrator study that had been prepared by Industrial
Surveys, a marketing research firm in Denver. He realized
that he needed to formulate some recommendations with
respect to some very specific questions. Should individual personal interviewers be used as suggested by Industrial Surveys, or should a series of one to six focus-group
interviews be used instead? Was the questionnaire satisfactory? Should individual questions be added, deleted,
or modified? Should the flow be changed? Exactly who
should be sampled, and what should the sample size be?
Research Setting
About 20 field salespeople at Mountain Bell Telephone
Company were involved in sales of communication
equipment and services to the health-care industry.
Because of job rotations and reorganizations, few salespeople had been in their present positions for more
than three years. They were expected to determine customer needs and problems and to design responsive
communication systems. In addition, there was a
health-care industry manager, Andy Smyth, who had
overall responsibility for the health-care industry marketing effort at Mountain Bell, although none of the
sales personnel reported directly to him. He prepared
a marketing action plan and worked to see that it was
implemented. The marketing action plan covered:
Sales objective by product and by segment
Sales training programs
Development of sales support materials and
information
Andy Smyth was appointed only recently to his
current position, although he had worked in the healthcare market for several years while with the Eastern
Bell Telephone Company. Thus, he did have some firsthand knowledge of customer concerns. Further, there
was an AT&T marketing plan for the health-care industry which included an industry profile; however, it
lacked the detailed information needed, especially at
the local level. It also lacked current information as to
competitive products and strategies.
Mountain Bell had long been a quasi-monopoly,
but during the past decade had seen vigorous aggressive competitors appear. Andy Smyth thought it imperative to learn exactly what competitive products were
making inroads, in what applications, and the basis of
their competitive appeal. He also felt the need for some
objective in-depth information as to how major Mountain Bell customers in the health-care industry perceived
the company’s product line and its sales force. He
hypothesized that the sales force was generally weak in
terms of understanding customers’ communication
needs and problems. He felt that such information
would be particularly helpful in understanding customers’ concerns and in developing an effective sales
training program. He hoped that the end result would
be to make the sales force more customer oriented and
to increase revenues from the health-care market.
209
While at Eastern Bell, Andy Smyth had initiated a
mail survey of hospital administrators that had been of
some value. Several months before, he had approached
Jim Martin with the idea of doing something similar at
Mountain. Jim’s reaction was that the questionnaire
previously used was too general (that is one question
was: What basic issues confront the health-care area?)
or too difficult to answer (How much do you budget
monthly for telecommunications equipment or service?
0–$1,000; $1,000–$2,000; etc.) Further, he felt that indepth individual interviews would be more fruitful.
Thus, he contacted Industrial Surveys, which, after considerable discussion with both Jim and Andy, created
the research design. They were guided by the following
research objectives:
1. What are the awareness and usage levels of
competitive telecommunications products by the
hospital?
2. What is the perception of Mountain Bell’s sales force
capabilities as compared to other telecommunications vendors?
3. What is the decision-making process as it pertains to
the identification, selection, and purchase of telecommunications equipment?
4. What concerns/problems impact most directly upon
the hospital’s (department’s) daily operations?
5. What are the perceived deficiencies and suggestions
for improvement of work/information flow?
Research Design
Research interviews will be conducted in seven Denver
area hospitals with the hospital administrator and,
where possible, with the financial officer and the
telecommunications manager. A total of 14 interviews
are planned. Interviews will be held by appointment,
and each respondent will be probed relative to those
questions that are most appropriate for his or her position and relevant to the study’s overall objectives. The
cost will be from $6,500 to $8,500, depending on the
time involved to complete the interviews. The questionnaire to be used follows.
A. Awareness and Usage of Competitive Telecommunications Equipment
1. What departments presently use non-Bell voice
communications equipment (paging, intercom,
message recording, etc.)? What were the main
considerations in selecting this equipment?
2. What departments use non-Bell data terminals
(CRTs)? What are the major functions/activities that this equipment is used for? What
were the main considerations in selecting this
equipment?
3. How do you view the capabilities of Bell System
voice-communications equipment to meet your
operations needs?
210
B.
C.
D.
E.
4. How do you view the capabilities of Bell System data terminals to meet your records- and
information-retrieval needs?
5. What do you feel are Mountain Bell’s main
strengths and/or weaknesses in meeting your
hospital’s overall telecommunications needs?
Perceptions of the Mountain Bell Sales Force
1. What should a telecommunications specialist
know about the hospital industry in order to adequately address your voice-communications and
data-processing needs?
2. Have you ever worked with any Mountain Bell
marketing people in terms of your communications needs? If so, how knowledgeable do you
perceive the Mountain Bell sales force to be with
respect to both the health-care industry and their
telecommunications equipment? How do they
compare to non-Bell vendors of such products?
Purchasing Decision
1. What is the standard procedure for selecting and
authorizing a telecommunications purchase? Is
this based primarily on the dollar amount involved or type of technology?
2. Who has the greatest input on the telecommunications decision (department manager, administrator, physicians, and so on)?
3. What are the most important considerations in
evaluating a potential telecommunications purchase (equipment price, cost-savings potential,
available budget, and so on)?
4. What supplier information is most important in
facilitating the purchasing decision? How effective has the Mountain Bell sales force been in providing such information?
Specification of the Most Important Problems or
Concerns Relating to Effective Hospital Management
1. What are the most important problems or concerns confronting you in managing the hospital?
2. What type of management data is required in order
to deal effectively with these problems or concerns?
3. How are these data presently recorded, updated,
and transmitted? How effective would you say
your current information-retrieval system is?
4. Do you have any dollar amount specifically budgeted for data or telecommunications improvements in 2004–2005? What specific information
or communication functions are you most interested in upgrading?
Achieving Maximum Utilization of Hospital Facilities
1. Do you experience any problems in obtaining
accurate, up-to-date information on the availability of bed space, operating rooms, or lab services?
2. Do you see _________ hospitals as competing
with other area hospitals or HMOs in the provision of health-care services? If so, with which
hospitals? Do you have a marketing plan to deal
with this situation?
F. Efficient Use of Labor Resources
1. How variable is the typical daily departmental
workload, and what factors most influence this
variance?
2. How do you document and forecast workload
fluctuations? Is this done for each hospital
department?
3. To what extent (if any) do you use outside consulting firms to work with you in improving the
delivery of hospital services?
G. Reimbursement and Cash Flow
1. Which insurer is the primary provider of funds?
How is reimbursement made by the major
insurers?
CASE 8-2
U.S. Department of Energy (A)
Judy Ryerson, the head of the windmill power section
of the U.S. Department of Energy, was considering
what types of qualitative marketing research would be
useful to address a host of research questions.
The U.S. Department of Energy was formed to deal
with the national energy problem. One of its goals was
to encourage the development of a variety of energy
sources, including the use of windmill power. One difficulty was that almost nothing was known about the
current use of windmill power and the public reaction
to it as a power source. Before developing windmill
power programs, it seemed prudent to address several
2. What information do you need to verify the existence and type of insurance coverage when an
individual is being processed for admission or
outpatient hospital services? What, if any, problems are experienced in the verification and communication of insurance information?
Source: This case was prepared by D. Aaker
and J. Seydel as a basis for class discussion.
research questions to obtain background information
and to formulate testable hypotheses.
Current Use of Windmills in the United States
How many power-generating windmills are there?
Who owns them? What power-generating performance
is being achieved? What designs are being used? What
applications are involved?
Public Reaction
What are the public attitudes to various power sources?
How much premium would the public be willing to pay
for windmill power sources, both in terms of money and
in terms of “visual pollution?” What is the relative acceptance of six different windmill designs ranging from the
“old Dutch windmill” design to an egg-beater design?
Assignment
ge / aak
w
om/ c o l
er
.c
le
w. wi l ey
Design one or more qualitative research designs to
address the search questions and to develop hypotheses for future testing. If focus-group interviews are
CASE 8-3
Acura
w
Acura was introduced into the U.S. luxury market in
1986 as a premium car at a bargain price, compared
with European luxury imports. Given a $20,000 price
tag, it became the best-selling premium import brand.
considered, provide a set of questions to guide the
moderator.
As long as it could underprice the competition,
Acura’s sales were steady. In 1992, however, European
luxury brands such as Mercedes, BMW, and Audi
started slashing their prices, and the value of the
Japanese yen rose. Both factors combined to wipe out
Acura’s “premium car at a bargain price” advantage.
Another blow came when archrivals Toyota Motor
211
Corp. and Nissan Motor Corp. launched the Lexus
and Infiniti, even more upscale brands, starting at
$35,000.
I. Campaign Time
Acura did not anticipate the resurgence of the European luxury brands. Acura is strong in quality,
dependability, and value, but it does not perform well
in the luxurious, expensive-looking, and prestige
departments. However, as Acura discovered, customers want performance plus status, prestige, comfort, and luxury. So far, Acura has struggled to justify
prices and convince buyers that the $40,000 maximum price on the Acura Legend will include those
attributes.
In the desire to “retool its image,” Acura replaced
its performance-based ads with luxury ads, such as a
campaign nicknamed “Chanel ads,” featuring mansions with cobblestone driveways and a towel-draped
woman getting a massage in a gazebo. The tag line
reads, “Some things are worth the price.” Competition
has rebutted by claiming that their $13,000 Altima was
“worth far more than the price.”
Other efforts by Acura to “retool” their image have
included promotional tours with Cirque du Soleil. Buyers
might be getting mixed signals. One-third of Acura’s
regional budget is allocated to dealerships, which typically run hard-sell rather than image ads. Acura has also
struggled to convince dealers to upgrade their showrooms to fit the luxury image. The line’s well-established
brand names, Legend and Integra, will soon be replaced
by bigger, more luxurious cars designed to compete with
top-of-the-line, $50,000-plus Lexus and Infiniti models.
Equipped with a smaller engine, RL model Acura cars
will sell for $10,000 less.
II. Issues
Luxury service and new luxury products will be required
to win customers. Should Acura keep pouring dollars into
the image campaign, or stick to performance and quality,
its strengths? What do the consumers want? Design a format for a focus-group discussion to analyze this situation.
(Hint: Consider three formats: exploratory, clinical,
experiencing. Will the prestige pitch work to rescue lagging sales?)
Source: This case is based on Larry Armstrong,
“Acura: Stuck between Gears?,” Business Week,
October 2, 1995, pp. 136–137, and was prepared by
V. Kumar for classroom discussion.
CASE 8-4
Exploratory Research on the
Lleyton Hewitt National Visa Card
tournament after his recovery, but he was defeated by
Alberto Martin in the early rounds. He attributed his
early loss to the fact that he was still recovering from
his illness and lacked energy.
The Player
Lleyton Hewitt, a top-class tennis player from Australia,
stunned the world by defeating the reigning champion
Pete Sampras to win the U.S. Open in 2001.
His inherent talent was obvious when he beat his
older and more experienced opponents at the tender
age of 8. He won his first professional title in 1998
when he was 17 years old. He decided to discontinue
his studies to pursue a full-time career in tennis. His
victories, over top-seeded players such as Patrick
Rafter, Pete Sampras, Gustavo Kuerten, Alex Corretja,
Magnus Norman, and many others during the Association of Tennis Professionals (ATP) tour, established him
as the number-one tennis player in the world.
Unfortunately, however, Hewitt had to withdraw
from the Hopman Cup in 2002 because he had chicken
pox. The 2002 Australian Open was his comeback
The Lleyton Hewitt Credit Card
In January 2002, Hewitt agreed to be featured on a new
Visa credit card offered by the National Australia Bank.
With the Lleyton Hewitt National Visa Card, the bank
expected to capitalize not only on the popularity of this
young player but also on the popularity of Australia’s
premier tennis event, the Australian Open. One of the
salient features of the card was a low introductory
interest rate of 6.9 percent p.a for the first six months.
Furthermore, cardholders could choose between the
Standard National Visa Card and the Gold National
Visa Card.
The extensive market research conducted by the
bank prior to the release of the credit card revealed a
strong demand for such a product. The research also
indicated that the association of Lleyton Hewitt with
the card would increase customer acquisition, card
212
usage, and customer retention. Target customers were
short-listed based on the information available about
their demographics and psychographics.
The National Australia Bank, which over the previous 18 months had aggressively pursued a strategy
of offering innovative products, believed that the
Lleyton Hewitt National Visa Card was an example of
an innovative lifestyle product.
Questions for Discussion
1.
2.
3.
What types of demographic and psychographic
information are required for this study? How
would you obtain it?
Would the recent poor performance of Lleyton
Hewitt affect the attitude of prospective and current credit card holders of the bank? How would
the company carry out marketing research to study
this issue?
Design the focus-group study and develop some
key questions that the interviewers can use to
elicit the level of demand for the credit card.
How would the demand for the credit card be
forecasted?
CASE 8-5
Hamilton Beach Conducts Primary
Research in Mexico and Europe
Hamilton Beach is well-known in the United States as
a maker of small kitchen appliances like blenders,
toasters, and mixers. However, the brand is not so
familiar in other countries. When the firm decided to
expand into Europe and Mexico with a new product
line, they wanted to assess the brand awareness in
these markets. Based on the assumption that the Mexican consumer is slightly more prone to buying U.S.
products, the firm wanted to confirm that there was
better brand awareness in Mexico.
A mixture of qualitative and quantitative research
techniques were used in Mexico. Through the study,
the firm was looking to determine what features were
to be included in the product and the retail environment through which they were going to sell. They also
wanted to find out the acceptance level for the brand
and whether the features of the product fit in well with
consumer needs. They found that the brand awareness
4.
How would Hewitt’s age, lifestyle, personal interests, and personality relate to those of the cardholders? How would you design a marketing
research study to identify the demographic characteristics of the prospective cardholders and identify the future customer group?
This case was prepared by the authors with inputs
from “Lleyton Hewitt Serves up a Winner for the
National,” press release, http://www.nab.com.au/
wps/wcm/connect/nab/nab/home/about_us/
8/5/21/1/2, January 10, 2002.
was high in Mexico, and considering the proximity to
the United States and the similar taste preferences due
to exposure to the same types of products, they were
optimistic in going ahead with the product launch. The
initial results were promising and indicated success.
Then Hamilton Beach changed its focus to the
European market.
The research in Europe involved focus groups with
the participants being asked to look at and touch the
products. After the participants had an opportunity to
examine the products, they were asked to come back
and talk about them in general, spontaneous terms. The
discussion showed that European consumers were
much more focused on design than American consumers. They seemed to like colors, rounded shapes,
and good-looking appliances, especially those they plan
to leave out on their kitchen counters. European consumers also did not appear to be as interested in product features. For example, while an American consumer
perceives an 18-speed blender as definitely added value,
the European consumer seemed to perceive it as confusing and therefore as something negative.
213
The research also highlighted the differences in how
Hamilton Beach product designers and consumers
define a product line. To the designers, uniformity of
things like the color of knobs on the appliances and the
fonts used in the instructions make a group of products
into a product line. But European consumers felt
product attributes like color and shape defined a product line. Hamilton Beach decided to change the whole
European line to suit the consumer preferences and go
back and do the same kind of research, before launching the product on a large scale.
Questions for Discussion
1.
Do you think the focus group research conducted
in Europe was representative of the population in
terms of:
■ Brand awareness and acceptance
■ Desired product features
■ Distribution channels
2.
As a market research analyst, is there an alternate
way to design the research process in Europe to
address the issues?
Source: Adapted with permission from
Joseph Rydholm, “Seeking the Right Mix,”
November 2001, http://www.quirks.com/articles/a2001/
20011105.aspx?searchID=218469119&sort=9
Mini Project
The mini project is a group activity designed for students wherein they are required to see video clips
from websites (for example, www.pbs.org and
www.leaduser.com). The topics of the clips include
narrowcasting, narrowing boundaries between content
and advertising, science of selling, and lead-users and
innovation. The students are then asked to present the
topic in class and be prepared to answer questions
related to the topics discussed. Grading for the presentation is based on the students’ ability to clearly
214
define the concept assigned, explanation of the concept
using relevant examples, and incorporation of their
relation to contemporary marketing research issues.
Students are also encouraged to refer to other relevant
sources in designing the content of the presentation.
This project is aimed at giving students a real-world
look at both qualitative marketing research and data
analytics issues. Additional information on this project
including the URLs for the topics is available at
http://www.drvkumar.com/MR11.
INFORMATION FROM RESPONDENTS:
ISSUES IN DATA COLLECTION
9
Learning Objectives
■ Discuss the different kinds of information that are collected through surveys.
■ Introduce the errors that occur while conducting a survey.
■ Discuss the various factors that influence the selection of the survey method to use.
■ Discuss the ethical issues involved in collecting data from respondents.
The survey is the overwhelming choice of researchers for collecting primary data. The methods
already discussed—qualitative and observational research, secondary data analysis, and secondary data research on the Internet—are more likely to be used to improve or supplement the survey method than to take its place.
The principal advantage of a survey is that it can collect a great deal of data about an individual respondent at one time. It is perhaps only stating the obvious to say that for most kinds
of data the respondent is the only, or the best, source. The second advantage of this method is
versatility; surveys can be employed in virtually any setting with any group—whether among
teenagers, old-age pensioners, or sailboat owners—and are adaptable to research objectives that
necessitate either a descriptive or causal design.
These advantages are not easy to achieve. Effective implementation requires considerable
judgment in the choice of a survey method, whether a personal or telephone interview or a selfadministered survey such as an email, mail or fax questionnaire. There also are some distinct
disadvantages to surveys that stem from the social interaction of interviewer and respondent.
Indeed, a survey cannot be developed or interpreted properly without a knowledge of the errors
that may bias the data during this interaction. These and other related issues are discussed in
this chapter as a prelude to the analysis of the methods of collecting survey data, the principal
focus of the next chapter.
INFORMATION FROM SURVEYS
Surveys can be designed to capture a wide variety of information on many diverse topics and
subjects. Attitudes are very often the subject of surveys. Information on attitudes frequently is
obtained in the form of consumers’ awareness, knowledge, or perceptions about the product, its
features, availability, and pricing, and various aspects of the marketing effort. Surveys can also
capture the respondents’ overall assessment and the extent to which the object is rated as favorable or unfavorable. Information can be obtained about a person’s image of something. Each person tends to see things a little differently from others, so no two images are apt to be exactly
alike. Decisions are also the topic of research, but the focus is not so much on the results of decisions in the past but more on the process by which respondents evaluate things. Those seeking
survey information are often keenly interested in those aspects of the decision process that people use to choose actions. Marketers are often concerned with why people behave as they do.
Most behavior is directed toward satisfying one or more human needs. Thus, the answer to the
215
CHAPTER 9 INFORMATION
FROM
RESPONDENTS: ISSUES
IN
DATA COLLECTION
question of “why” is often obtained by measuring the relationship between actions and needs,
desires, preferences, motives, and goals.
Measuring behavior usually involves four related concepts: what the respondents did or did
not do; where the action takes place; the timing, including past, present, and future; and the frequency or persistence of behavior. In other words, it often means assessing what, where, when,
and how often. Surveys can also be conducted to determine respondents’ lifestyles. Groupings of
the population by lifestyle can be used to identify an audience, constituency, target market, or
other collections of interest to the sponsor. Social contact and interaction are often the focus of
survey research or bear heavily on other issues relevant to the survey. So the family setting,
memberships, social contacts, reference groups, and communications of respondents frequently
are measured or assessed within the survey research process. Demographic factors often
obtained through surveys include such variables as age, sex, marital status, education, employment, and income, among others. Personality reflects consistent, enduring patterns of behavior,
and it is more deeply rooted than lifestyle. Personality can be measured using rating methods,
situational tests, projective techniques, and inventory schemes. Motivation and knowledge are
also frequently measured using surveys.
SOURCES OF SURVEY ERROR
The process by which respondents are questioned appears deceptively simple. The reality, however, is closer to Oppenheim’s opinion that “questioning people is more like trying to catch a
particularly elusive fish, by hopefully casting different kinds of bait at different depths, without
knowing what is going on beneath the surface.”1
The problem of getting meaningful results from the interview process stems from the need
to satisfy reasonably the following conditions:
■ Population has been defined correctly.
■ Sample is representative of the population.
■ Respondents selected to be interviewed are available and willing to cooperate.
■ Respondents understand the questions.
■ Respondents have the knowledge, opinions, attitudes, or facts required.
■ Interviewer understands and records the responses correctly.
These conditions often are not satisfied because of interviewer error, ambiguous interpretation
of both questions and answers, and errors in formulating responses. These types of errors are
shown in Figure 9-1 as filters or screens that distort both the question and response. Ambiguity,
usually a consequence of poor question wording, is covered in Chapter 12. In this section, we
will deal with the factors that influence prospective respondents’ willingness to cooperate and
provide accurate answers to whatever questions are asked.
Nonresponse Due to Refusals
Refusal rates are highly variable. They can be as low as 3 to 5 percent of those contacted for a
short interview on a street corner or at a bus stop, to 80 or 90 percent and higher for lengthy
personal and telephone interviews or mail questionnaires, which are of little interest to most subjects. High refusal rates can be a major source of error if those who refuse to be interviewed are
very different from those who cooperate. People cooperate for a number of reasons, such as trying to be friendly and helpful, to interact socially when they are lonely or bored, to know more
216
SOURCES
Ambiguity of
question
(see Chapter 12)
POPULATION
FRAME
Sample
RESPONDENT
Nonresponse
due to refusals
or not-at-home
Sampling error
(see Chapters
14 and 15)
SURVEY ERROR
Interviewer
error
Question
Answer
• Inaccuracy in
response
• Inability to
formulate a
response
• Unwillingness
to respond
OF
INTERVIEWER
Ambiguity of
answer
FIGURE 9-1
Sources of error in information from respondents.
about the subject, or to experience something novel. Many people also cooperate when they
expect a reward or a direct benefit.
People refuse to answer survey questions for a number of reasons. Fear is the main reason for
refusal.2 Othes might be that some may think of surveys as an invasion of privacy; the subject matter
may be sensitive (such as death, sexual habits, etc.); or there may be hostility toward the sponsor.3
Refusal rates can be reduced if the reason for refusals is related to a phenotypic source. A phenotypic
source of refusals refers to characteristics of the data collection procedure (which questions are
asked, how they are asked, the length of the interview, and so on). These elements vary from study
to study and can be controlled to a certain extent. But if the source of refusals is genotypic, then it
is difficult to control for them. A genotypic source refers to the indigenous characteristics of respondents (such as age, sex, and occupation). One avenue to reducing this type of refusal is to try to
interview respondents in ways that are less intrusive in their lives.4 Other methods, such as multiple callbacks (mailings), guarantee of anonymity, increasing surveyor credibility, incentives for completed responses, randomized responses, and shortening the survey through matrix designs, can
increase response rates.5 An example of the effect of nonresponse errors is given in Marketing
Research in Action 9-1.
MARKETING RESEARCH IN ACTION 9-1
Nonresponse Bias Leads to Wrong Prediction in Presidential
Opinion Poll
An example of significant bias as a result of low response to mail questionnaires is the often-cited Literary
Digest presidential poll in 1936. The two candidates were Alfred E. Landon of the Republican Party and
Franklin D. Roosevelt of the Democratic Party. The Literary Digest conducted a mail survey and predicted
a victory for Landon in the election. But Roosevelt won by a big margin in a landslide victory for the
Democrats. The story is told that the sample was drawn from telephone books, and the Republicans
(Landon’s party) were more likely to have telephones in 1936. Hence, this biased the sample toward
Republicans. In addition, the failure of the poll to predict correctly was attributed to nonresponse; only
a minority of those asked to return questionnaires did so. As is typical of mail surveys, those who
returned the questionnaire wanted the underdog Landon to win. The Landon supporters were particularly likely to want to express their views, whereas the Roosevelt supporters, since they were the majority,
did not bother to respond to the survey.
SOURCE: M. Bryson, “The Literary Digest Poll: Making of a Statistical Myth,” American Statistician, November 1976, pp. 184–185.
217
CHAPTER 9 INFORMATION
FROM
RESPONDENTS: ISSUES
IN
DATA COLLECTION
Inaccuracy in Response
Respondents may be unable to give any response, or unwilling to give a complete and accurate
response.
Inability to Respond. Respondents may not know the answer to a question because of ignorance or forgetting, or may be unable to articulate it. All three problems create further errors
when respondents contrive an answer because they do not want to admit that they don’t know
the answer or because they want to please the interviewer. Respondents are likely to be ignorant when asked some questions. For example, a husband and wife may not be equally aware
of the financial status of the family (such as insurance, investment, and benefits), so depending
on who you survey you will get a different answer.
The likelihood of forgetting an episode such as a visit to a doctor, a purchase, and so forth,
depends on both the recency of the occurrence and the importance of the event, as well as on
what else was happening at the same time. 6 Ideally, questions should be asked only about recent
behavior. If retrospective questions are required, the accuracy of recall can be improved by questioning the respondent about the context in which an event occurred. Memories may be sharpened by aided-recall techniques. These attempt to stimulate recall with specific cues, such as
copies of magazines, pictures, or lists. It is always preferable to ask about specific occasions of
an activity.
It is essential to keep in mind that most respondents want to be cooperative; when in doubt,
they prefer to give too much rather than too little information. Memory distortions compound this
source of error by telescoping time, so that an event is remembered as occurring more recently
than it actually did.7 Another common memory error is averaging, whereby something is
reported as more like the usual, the expected, or the norm. This is a particular problem for
researchers trying to study the exceptions to the ordinary. General Foods’ researchers interested
in homemaker’s variations in evening meals found it very difficult to overcome the respondent’s
tendency to say “It was Sunday, so it must have been roast beef.” In studies of evening meals
at General Foods, researchers reduced the averaging problem by asking respondents a series of
questions that helped them to reconstruct the event. A third memory pitfall is called omission,
where a respondent leaves out an event or some aspect of an experience. Respondent fatigue
and poor interviewer rapport are demotivating and result in increased omitting. The use of
graphic aids to recall or recognition measures will help to reduce omission.
Finally, respondents may be unable to respond because they cannot formulate an adequate
answer. This is especially true of direct questions about motivations. Many choice decisions are
made without conscious consideration of the reasons. So when people are asked why they
responded to a charitable appeal, bought a particular brand of analgesic, or watched a certain
television program, the reasons they give are likely to be incomplete and superficial. An alternative is to use indirect methods such as the projective techniques described in Chapter 8.
One technique that is being used to judge the probability of response inaccuracies is to measure the time it takes for the respondent to answer a question. Studies show that a too-slow
response may mean the question is too difficult to understand, whereas an immediate response
probably means that it was either misunderstood or not well thought out.
Unwillingness to Respond Accurately. During the interview a number of response bias factors
may come into play to subvert the positive motivations that were present when the respondent
agreed to participate. Questionnaires that are lengthy and boring are especially vulnerable to
these biases.
218
1. Concern about invasion of privacy. Although most respondents don’t regard a survey, per se, as an
invasion of privacy, their tolerance may not extend to detailed personal questions. As many as
20 percent of the respondents in a telephone or personal interview survey may refuse to answer
an income question, and others may distort their answer. To some extent, assurances of confidentiality and full explanations of the need for the data will reduce this problem.
SOURCES
OF
SURVEY ERROR
2. Time pressure and fatigue. As a lengthy interview proceeds, the accuracy of responses is bound
to decline. Those respondents who initially were reluctant to participate because they were
busy become anxious about the time that seems to be required. Not surprisingly, they may
decide that giving abrupt answers and avoiding requests for clarification are the best ways to
terminate the interview quickly. Even those who respond willingly and fully to all questions
will become fatigued eventually if the interview is too long. The resulting bias is most likely
random, but is sometimes in a consistent direction if the respondent decides to retaliate by
grouping all answers about some point on a scale.8
3. Prestige seeking and social desirability response bias. There is mounting evidence that respondents
will distort their answers in ways that (they believe) will enhance their prestige in the eyes of
the interviewer and will not put them at variance with their perception of the prevailing norms
of society. Consequently, questions that have implications for prestige—such as income,
education, time spent reading newspapers, tastes in food, or even place of residence—may be
biased subtly in ways that reflect well on the respondent.
4. Courtesy bias. There is a general tendency to limit answers to pleasantries that will cause little
discomfort or embarrassment to the interviewer, or to avoid appearing uncooperative. We have
seen already how this bias can inflate responses to aided-recall questions. It also is encountered
in concept-testing situations when a respondent gives a “courtesy endorsement” to the description of a new idea, even though he or she may have little interest in the idea.9
5. Uninformed response error. Simply asking someone a question implies that the interviewer
expects the respondent to have an answer. This expectation, plus a desire to appear cooperative, may induce respondents to answer a question despite a complete lack of knowledge
about the topic. When a sample of the general public was asked a question about a fictitious organization, the National Bureau of Consumer Complaints, an astonishing 75 percent
of those who returned the questionnaire expressed an opinion about the effectiveness of the
organization in obtaining relief for consumers.10
6. Response style. Evaluative questions requiring a good-bad, positive-negative judgment are
afflicted by systematic tendencies of certain respondents to select particular styles or categories of response regardless of the content of the question. For example, there is an acquiescence response set, which is the tendency to favor affirmative over negative responses. This
is different from “yea saying,” which is a tendency to give exaggerated responses; that is,
good becomes very good and bad becomes very bad. For example, Wells found that yeasayers consistently gave higher ratings to favorably evaluated advertisements and were
more likely to exaggerate self-reports of product purchases.11
Interviewer Error
Interviewers vary enormously in personal characteristics, amount of previous experience, style of
interviewing, and motivation to do a thorough job. The differences among interviewers also mean
a great deal of variability in the way interviews are conducted.
Respondent’s Impression of the Interviewer. For most respondents a personal interview is a sufficiently novel experience that the interviewer becomes a major source of clues to appropriate
behavior. The interviewer must be seen as a person who is capable of understanding the respondent’s point of view, and of doing so without rejecting his or her opinion. This kind of rapport is
most likely to be established quickly when respondent and interviewer share basic characteristics
such as sex, age, race, and social class.
The attitudes the interviewer reveals to the respondent during the interview can greatly affect
their level of interest and willingness to answer openly. Especially important is a sense of assurance and ease with the task. Communication will be inhibited further if the interviewer appears
flippant or bored, constantly interrupts the person when speaking, or is too immersed in note
taking to look up.12 Obviously, proper selection of interviewers, coupled with good training, can
reduce many of these problems.
219
CHAPTER 9 INFORMATION
FROM
RESPONDENTS: ISSUES
IN
DATA COLLECTION
Questioning, Probing, and Recording. The way an interviewer asks a question and follows up
by probing for further details and clarification will be colored by (1) the interviewer’s own feelings about the “appropriate” answer to the questions and (2) expectations about the kind of
answers that “fit” the respondent. For example, when interviewing a person with limited education, the interviewer might shift unconsciously from a question worded, “Have any of your children attended college?” to “I don’t imagine any of your children have gone to college. Have they?”
In one study it was found that one interviewer obtained 8 percent while another interviewer
obtained 92 percent choosing the same option.13
Perhaps the most common interviewer error is insufficient probing. The respondents may
not be expected to have much to say about the subject or may have given an answer that the
interviewer thinks is “right.”
Fraud and Deceit. Modest interviewer compensation and the problem of monitoring the activities of personal interviewers out in the field, or telephone interviewers calling from their home,
provide ample incentive for cheating. This may be as serious as outright fabrication of an entire
interview or judicious filling in of certain information that was not obtained during the interview. Because it is such a serious potential source of error, most commercial research firms validate 10 to 15 percent of the completed interviews. This entails interviewing a sample of those
who were reported to have been interviewed, to verify that an interview actually took place and
that the questions were asked.
Improving Interviewer Quality. New approaches to improving data collection quality are being
constantly tried in order to overcome some of the interview problems that have been mentioned.
Field briefings can be improved by preparing videotaped briefings to show to interviewers, to
ensure a consistent message to all interviewers prior to data collection. Actual interviews can be
recorded by camera at a particular site, and all interviewers can be rotated through that site to
check on the quality of interviewing. Finally, if electronic quality checks are too expensive, research
firms can hire independent field personnel to check quality.
The array of problems and sources of error, summarized in Table 9-1, have the greatest effect
on the personal interview method. The best solution by far is to minimize the problems by proper
TABLE 9-1
Sources of Error in Interview Surveys
1. Nonresponse errors due to refusals
a. Fear of the consequences of participation
b. Resentment of an invasion of privacy
c. Anxiety about the subject
2. Inaccuracy in responses
a. Inability to give a response
i. Don’t know the answer
ii. Memory problems
iii. Problems in formulating an answer
b. Unwillingness to respond accurately
i. Concern about invasion of privacy
ii. Time pressure and fatigue
iii. Desire to enhance prestige
iv. Desire to appear cooperative
v. Biased response style
vi. Interviewer forcing the respondent to have an answer
220
3. Errors caused by interviewers
a. Provision of clues to “appropriate” responses
b. Inadequate questioning and probing
c. Fraud and deceit
FACTORS AFFECTING
THE
CHOICE
OF
A SURVEY METHOD
TABLE 9-2
Basic Survey Methods and Their Characteristics
Survey Method
Characteristics
Personal interviews
The interviewer interviews the respondent in person.
There is direct contact between the interviewer and the respondent.
The environment (mood of the respondent and the interviewer, the time
and place of the interview, etc.) affects the data collection process to a
large extent.
Costliest, and the most time-consuming form of data collection.
Telephone interviews
The interviewer interviews the respondent over the telephone.
The interviewer has only verbal contact with the respondent.
The environment plays a relatively minor role in the data collection
process.
Data collection cost is in between that of a personal interview and a
mail survey.
Self-administered surveys
(mail, email, web-based,
fax)
The questionnaire is completed by the respondent.
The interviewer has no contact with the respondent.
The environment plays no role in the data collection process.
The least expensive form of data collection.
recruiting, selecting, training, motivation, and control of interviewers. Yet the potential for poorquality interviews is enormous.14
While other methods of data collection are less prone to interview error, they have offsetting
problems, as we will see in Chapter 10.
METHODS OF DATA COLLECTION
The choice of a data collection method is a critical point in the research process. The decision is
seldom easy, for there are many factors to be considered and many variations of the three basic
survey methods: (1) personal interview, (2) telephone interview, and (3) self-administered survey (e.g. mail, email, fax, etc.). In this section, we will look briefly at the different methods of
data collection and the factors affecting the choice of method.
Because each research problem will have a different ranking of importance, and no data collection method is consistently superior, few generalizations can be made. Much depends on the
researcher’s skill in adapting the method to the circumstances. Overall, however, the telephone
and the self-administered survey methods are the dominant methods for conducting surveys. In
the 1990 Walker Industry Image Study, it was found that 69 percent of the respondents had participated in mail surveys, 68 percent had participated in telephone surveys, 32 percent in mall
intercept surveys, and 15 percent in door-to-door interviews. However, both in academic and
business environment, email and web-based surveys are being used increasingly. The characteristics of each survey method are explained briefly in Table 9-2.
FACTORS AFFECTING THE CHOICE
OF A SURVEY METHOD
One of the most important decisions a researcher must make is the way in which the data will
be collected. The decision to choose among the various survey methods already discussed is
affected by a number of factors.15 Some of them are described in the paragraphs that follow.
221
CHAPTER 9 INFORMATION
FROM
RESPONDENTS: ISSUES
IN
DATA COLLECTION
Sampling
The way a researcher plans to draw a sample is related to the best way to collect the data. Certain
kinds of sampling approaches make it easier or more difficult to use one or another data collection
strategy. If one is sampling from a list, the information on the list matters. Obviously, if a list lacks
either good mailing or email addresses or good telephone numbers, trying to collect data by mail,
phone, or mail/email is complicated. Random-digit dialing has improved the potential of telephone
data collection strategies by giving every household with a telephone a chance to be selected.
Of course, it is possible to use random-digit dialing strategies, which are explained in
Chapter 10, merely to sample and make initial contact with households, followed by collecting data using some other mode. Once a household has been reached, one can ask for a physical or email address, to permit either a mail or email questionnaire to be sent or an interviewer
to visit. Such designs are particularly useful when one is looking for a rare population, because
both the sampling and the screening via telephone are relatively less expensive than doing the
same task with a personal interviewer.
Another sampling issue to consider is designating a respondent. If the sample frame is a list
of individuals, any procedure, including mail, is feasible. Many surveys, however, entail designating a specific respondent at the time of data collection. If a questionnaire is mailed to a household or organization, the researcher has little control over who actually completes it. Therefore,
the involvement of an interviewer is a critical aid if respondent designation is an issue.
Type of Population
The reading and writing skills of the population and its motivation to cooperate are two salient
considerations in choosing a mode of data collection. Self-administered approaches to data collection place more of a burden on the respondent’s reading and writing skills than do interviewer
procedures. Respondents who are not very well educated, whose reading and writing skills in
English are less than facile (but who can speak English), people who do not see well, and people who are somewhat ill or tire easily will find an interviewer-administered survey easier than
filling out a self-administered form. Another problem for self-administered surveys is getting
people to return a completed questionnaire. People who are particularly interested in the research
problem tend to be most likely to return self-administered questionnaires.16
Question Form
Generally speaking, if one is going to use one of the self-administered questionnaire methods, it
is best to use closed-end questions—that is, questions that can be answered by simply checking
a box or circling the proper response from a set provided by the researcher. Second, and more
important, self-administered open-ended questions often do not produce useful data. With no
interviewer present to probe incomplete answers for clarity and to meet consistent question objectives, the answers are not often comparable across respondents, and they will be difficult to code.
There are question forms (including those with complex descriptions of situations or events
and those requiring pictures or other visual cues) that cannot be adapted to the telephone. If
such measurement is a critical part of a survey, some form other than the telephone survey probably is needed.
Question Content
222
Researchers have argued persuasively that one or another of the strategies should have an advantage when dealing with sensitive topics. Self-administered surveys are thought to be best,
because the respondent does not have to admit directly to an interviewer a socially undesirable
FACTORS AFFECTING
THE
CHOICE
OF
A SURVEY METHOD
or negatively valued characteristic or behavior. Others have argued that telephone surveys lend
an air of impersonality to the interview process that should help people report negative events
or behaviors. Moreover, random-digit dialing at least provides the option of having a virtually
anonymous survey procedure, because the interviewer need not know the name or location of the
respondent. Still others argue that personal interviews are the best way to ask sensitive questions, because interviewers have an opportunity to build rapport and establish the kind of trust
that is needed for respondents to report potentially sensitive information.
An entirely different aspect of question content that may affect the mode of data collection
is the difficulty of the reporting task. In some surveys, researchers want to ask about events or
behaviors that are difficult to report with accuracy, because they extend over a period of time or
are quite detailed. In such cases, reporting accuracy may benefit from a chance to consult records
or to discuss the questions with other family members. The standard interview is a quick questionand-answer process that provides little such opportunity; this is especially true for telephone
interviews. Self-administered procedures provide more time for thought, for checking records,
and for consulting with other family members.
Response Rates
The rate of response is likely to be much more salient than other considerations in the selection
of a data collection procedure. Obviously, one of the strengths of group-administered surveys,
when they are feasible, is the high rate of response. Generally speaking, when students in classrooms or workers in job settings are asked to complete questionnaires, the rate of response can
approach 100 percent.
There is no doubt that the problem of nonresponse is central to the use of self-administered
surveys. If one simply mails questionnaires to a general population sample without an appropriate follow-up procedure, the rate of return is likely to be less than 10 percent. In cases of short
and well-formatted questionnaires, email surveys can be an efficient way to collect information
in a short period of time. It would be relatively easy for a participant to fill in the questionnaire,
and email it back to the surveyor, resulting in quicker responses than mail surveys. It would also
be economical to email reminders, which would help increase response rates. However, fax surveys would perform better when Internet penetration is low.
The effectiveness of telephone strategies in producing high response rates depends in part
on the sampling scheme. Response rates in some urban areas benefit from using the telephone,
whereas suburban and rural rates are usually lower for telephone surveys than when a personal interviewer is used. Using the telephone permits better coverage of units in buildings with
security systems and neighborhoods where interviewers are reluctant to go in the evening.
As discussed earlier, the issue of nonresponse bias may appear more pertinent for surveys
with low response rates. This is especially true in the international context. The key to addressing the nonresponse bias is in ensuring the representativeness of the responding sample. Even
surveys with relatively high response rates (e.g., 70–80%) may be just as susceptible to nonresponse bias as those with relatively low response rates (e.g., 30–40%). Therefore, understanding the nature of potential nonresponse bias should be considered to be of importance while
conducting surveys. Table 9-3 provides the response rates for various countries.
Costs
The great appeal of mail or email and telephone survey procedures is that in most cases they
cost less than personal interviews. Survey costs depend on a multitude of factors. Some of the
more salient factors are the amount of professional time required to design the questionnaire,
the questionnaire length, the geographic dispersion of the sample, the availability and interest
of the sample, the callback procedures, respondent selection rules, and the availability of trained
staff. Although on the surface mail survey costs might appear to be lowest, the cost of postage,
of clerical time for mailing, and of printing questionnaires turns out not to be trivial. Moreover,
223
CHAPTER 9 INFORMATION
FROM
RESPONDENTS: ISSUES
IN
DATA COLLECTION
TABLE 9-3
Response Rates for a Sample of Countries
Country
Survey & Response Rate (effective sample size)
United Kingdom: Confederation
of British Industry
Industry 45% (2000), Retail 22% (1000), Investment 19% (2000),
Services 20% (1000).
South Africa: Bureau for
Economic Research
40–45% across retail/wholesale, manufacturing, motor trade and
contractors (building related). Total sample of around 3000.
Canada: Statistics Bureau
Business conditions for manufacturing (55%).
Finland: Statistics Bureau
Consumer Survey 74% (2,200).
Austria: WIFO
Manufacturing 30%, Construction 30%, Services 34%. These are
panel response rates; response rates of recruits are lower.
France: INSEE
Industry 81% (4000), Investment 77% (4000), Retail 82% (5,000),
Services 76% (4500), Construction 84% (5,000).
Netherlands: Statistics Office
Industry 90%.
Slovenia: Statistics Office
Consumer Survey 68% (1500), Industry 91% (600), Construction
92% (300), Services 92% (520), Retail 82% (800).
Japan: Central Bank
Whole economy: TANKAN survey 95%.
Germany: IFO
Manufacturing 90%, Construction 70%, Overall including other
sectors 80%.
Slovak Republic: Statistics Office
Industry 75% (500), Construction 85% (420), Retail 60% (420),
Services 63% (500).
Statistics Norway
Business Tendency Survey for Manufacturing, Mining and Quarrying 85% (700).
China
Whole economy 95%.
SOURCE: OECD Report on “Assessing and minimising the impact of non-response on survey estimates,” OECD Statistics Directorate, November 2005.
if there are telephone follow-ups, the expense gets higher. Email and web-based surveys can cost
the least since no long-distance telephone charges are present and there are minimal administrative costs. Further, there is no need to reproduce the questionnaires.
Available Facilities
224
The facilities and staff availability should be considered in choosing a data collection mode.
Developing an interviewing staff is costly and difficult. Attrition rates are generally high for
newly trained interviewers. Many new interviewers are not very good at enlisting the cooperation of respondents, producing high refusal rates at the start. In addition, people who are good
at training and supervising interviewers are not easy to find. Thus, one very practical consideration for anyone thinking about doing an interviewer-conducted survey is the ability to execute a
professional data collection effort.
The available facilities are a crucial factor in the choice of a survey method in the international context. The question that has to be evaluated is whether local researchers must be used
in conducting the survey or whether to use foreign researchers. In considering whether to use
local researchers or import foreign researchers with specific skills, a number of factors need to
be evaluated. This question arises particularly in the context of developing countries, where
research staff with specific skills, such as the ability to conduct in-depth interviews or focus
groups, or skills in research design and analysis, may not be readily available.
The major advantage of using local researchers for qualitative research or in research
design is that they will know the local culture and people, as well as the language, and hence
be best able to understand local cultural differences.
On the other hand, local researchers may have more limited research experience than foreign staff
and may not have the same specialist skills. In interviewing, they may have difficulties in adopting
ETHICAL ISSUES
IN
DATA COLLECTION
a neutral or objective stance relative to respondents, or may lack familiarity with the design of a
sophisticated research instrument. In some cases, there may be a scarcity of local researchers with
even the minimum required skills, thus necessitating consideration of importing foreign researchers.
A second issue is that of training field interviewers. This is particularly critical in cases where
a pool of experienced interviewers is not readily available. In political and social surveys conducted in developing countries, extensive training programs have been developed for interviewing. Upscale individuals in leadership positions, such as village headmen, teachers, and
prefects, have been found to be good interviewers in these situations.
Duration of Data Collection
The time involved in data collection varies by mode. Mail surveys usually take two months to
complete. A normal sequence involves mailing the questionnaires and waiting for the responses.
If the response rate is poor, then some more questionnaires are mailed and, finally, some telephone
or in-person follow-up is done if the second wave of mailing also does not produce a good response
rate. At the other extreme, it is quite feasible to do telephone or email surveys in a few days.
Surveys done in a very short period of time pay a cost in nonresponse, because some people cannot be reached during any short period. However, telephone and email surveys can be done more
quickly than mail or personal interview surveys of comparable size.
ETHICAL ISSUES IN DATA COLLECTION
Survey research is an objective process and therefore has explicit guidelines and principles. However, it takes place in a subjective context, and thus is vulnerable to distortion by its producers and
perceivers. Misrepresentation of the research itself through the use of inadequate sampling procedures or volunteers to obtain so-called “survey” information will be discussed in Chapter 14.
Misrepresentation of the data collection process stems from two principal sources. The first is
the representation of a marketing activity other than research, as research. The second is the abuse
of respondents’ rights during the data collection process under the rationale of providing betterquality research.
Consumers expect to be sold and to be surveyed, and they expect to be able to tell the
difference without great difficulty. When a selling or marketing activity uses the forms and language of survey research in order to mask the real nature of the activity being performed, it violates the public trust. Classic examples of this type of practice are the following:
1. The use of survey techniques for selling purposes. In this case, a person answers a few questions only to find him/herself suddenly eligible to buy a specific product or service. The
misuse of the survey approach as a disguise for sales canvassing is a widespread practice
that shows no signs of abating.
2. The use of survey techniques to obtain names and addresses of prospects for direct marketing. These
efforts are usually conducted by mail. Questionnaires about products or brands are sent to
households, and response is encouraged by the offer of free product samples to respondents.
The listing firms compile the information by implying to the prospective customer that he or
she has been interviewed in a market study.17
In both of these cases, the consumer is not told that the company conducting the survey is in a business other than research. The harm caused by this practice is that legitimate research is given a bad
name in the eyes of consumers. Both response rates and response quality are jeopardized.
Even companies that practice legitimate research can violate the rights of respondents by
deliberately engaging in a number of practices such as:
■ Disguising the purpose of a particular measurement, such as a free draw or free product choice
question
225
CHAPTER 9 INFORMATION
FROM
RESPONDENTS: ISSUES
IN
DATA COLLECTION
■ Deceiving the prospective respondent as to the true duration of the interview
■ Misrepresenting the compensation in order to gain cooperation
■ Not mentioning to the respondent that a follow-up interview will be made
■ Using projective tests and unobtrusive measures to circumvent the need for a respondent’s
consent
■ Using hidden tape recorders to record personal interviews (or recording phone conversations
without the respondent’s permission)
■ Conducting simulated product tests in which the identical product is tried by the respondent
except for variations in characteristics, such as color, that have no influence on the quality of
a product
■ Not debriefing the respondent18
Many of these practices cannot be condoned under any circumstances, and others present the
conscientious researcher with a serious dilemma. Under certain circumstances, disguising the
nature of the research hypotheses may be the only feasible method by which to collect the necessary data. Yet these practices have the potential to create biased data and suspicion or resentment that later may be manifested in a refusal to participate in subsequent studies.
Both the misrepresentation of research and the abuse of respondents’ rights during the
legitimate research interviewing process involve consumer deception. This has two implications. From a business perspective, if public willingness to cooperate with the research process
is adversely affected, the long-term statistical reliability of marketing research is jeopardized.
From a social perspective, consumer deception violates basic business ethics. The responsibility of business to society rests on a fundamental concern for the advancement of professional business practices. Marketing research depends on mutual trust and honesty between
the business community and society. Deception undermines the trust by using people as mere
instruments to accomplish unstated purposes.
Other ethical issues in data collection arise even when the respondent is not being deceived.
One of the most important is invasion of privacy. A basic criterion of good research is respect
for the privacy of the individual, yet legitimate research is always to some extent intrusive in
nature. Excessive interviewing in certain metropolitan areas, especially in those popular for test
markets, can be perceived by consumers as violating their basic right to privacy. A recent telephone survey with 300 respondents found that 57 percent had been interviewed at some time
in the past and 50 percent had been interviewed during the past year. This type of overinterviewing may result in consumers feeling resentful and less willing to cooperate fully. If they
suspect that participation in one interview leads to further requests for interviews, they will be
more likely to refuse.
Ethics in Online Data Collection
226
With more people taking to the Internet, online data collection has become an attractive
option for all types of research. People flock to “chat rooms” and message boards to share
experiences or thoughts or ideas with “virtual” strangers, thus making an abundance of
social and behavioral information available on the ’Net. The Internet thus offers an ideal
medium for collecting data from widely dispersed populations at relatively low cost and in
less time than similar efforts in the physical world.19 While offering the comfort of data
collection from a single physical location, the Internet does raise concerns about the ethical
and legal dimensions of such research. Even when respondents choose to remain anonymous,
privacy issues and the unclear distinction between public and private domains make it possible for researchers to record their online interactions without the knowledge or consent of
the respondents.
QUESTIONS
AND
PROBLEMS
This is all the more important when it comes to social and behavioral research where the confidentiality of the respondents may not be maintained. For instance, in a study about sexually
abused survivors,20 researchers had collected and analyzed information from an online support
group, on condition of anonymity. Though the published report did not disclose the names of the
participants, it quoted the text of the participant’s postings as they were, making it possible for
the respondents to identify themselves. Also for a determined reader of the report, it would be
easy to trace the message back to the participant, if the name of the studied group were known.
Thus, while the Internet does offer logistical comforts in data collection, it raises serious questions about interpretation and applicability guidelines governing online data collection.
Summary
There are bound to be a number of errors when surveys are conducted, some of which can be controlled
and others of which cannot be controlled. A researcher should know all the potential sources of errors and
should try to reduce their impact on the survey findings. This is a good research practice and will lead to
more robust results.
The choice of a survey method, that is, whether to use self-administered, or telephone or personal interviews, is determined by a number of factors. A knowledge of these factors will make it easier for the researcher
to decide among the various methods. The most important factors are the sampling plan to be employed, the
type of population to be surveyed, the response rates required, the budget for the survey, and the available
resources. A number of factors may be important for the choice of survey method other than those discussed
here. It depends on the particular study, but the factors listed in the text are the most common ones.
All data collection methods are susceptible to misrepresentation by the researchers. Two principal
sources of this misrepresentation are disguising the true purpose of a marketing activity, such as selling,
by calling it marketing research, or abusing respondents’ rights during a legitimate data collection process.
Both of these practices involve a disregard of professional business ethics. Maintaining high standards of
business ethics while collecting data is the obligation of all responsible researchers.
Questions and Problems
1. How would you overcome some of the problems
you might anticipate in designing a survey to
establish the kind of paint used by the “do-ityourself” market when the members of this sample last redecorated a room?
2. How would you balance the requirements of improving the quality of interviews against any ethical
considerations that may arise?
3. Is the biasing effect of an interviewer more serious
in a personal or telephone interview? What steps
can be taken to minimize this biasing effect in these
two types of interviews?
4. People tend to respond to surveys dealing with
topics that interest them. How would you exploit
this fact to increase the response rate in a survey
of attitudes toward the local urban transit system,
in a city where the vast majority of people drive to
work or to shop?
5. If you were a marketing research manager, would
you permit the following if they were important to
the usefulness of a study?
a. Telling the respondent the interview would take
only two or three minutes when it usually took
four minutes and a follow-up 10-minute interview was employed.
b. Telling the respondent the questionnaire
would be anonymous but coding it so that the
respondent could be identified (so that additional available information about the respondent’s neighborhood could be used).
c. Secretly recording (or videotaping) a focusgroup interview.
d. Saying the research was being conducted by a
research firm instead of your own company.
6. You are product manager for Brand M butter,
a nationally known brand. Brand M has been
declining in absolute level of sales for the last four
consecutive months. What information, if any, that
could be obtained from respondents would be useful
for determining the cause or causes of this decline?
7. What are the general advantages and disadvantages associated with obtaining information by
questioning and by observation? Which method
provides more control over the sample?
8. Innovative Marketing Consultant has been contracted to conduct a job-satisfaction survey among
the 2,470 employees of United Machine Tools, Inc.,
Dayton, Ohio. The management insists that the
questionnaire be comprehensive and incorporate
all the views of the respondent in order to improve
the work environment.
a. What factors should Innovative Marketing Consultant consider in choosing the survey method
to be used?
227
b. Which question form would be most likely to
supply management with the information it has
requested?
c. Design a sample questionnaire that Innovative
Marketing Consultant will use in conducting
the job-satisfaction survey.
d. What possible sources of survey error might be
encountered?
e. United Machine Tools, Inc., has a wholly owned
subsidiary in Madras, India. What additional factors must Innovative Marketing Consultant consider in choosing the survey method to assess the
job-satisfaction levels of the employees at the
Madras plant?
9. Mark Hirst, a high school student in Sydney,
Australia, developed an innovatively shaped
surfboard that radically increased the number of
maneuvers a surfer could perform in low surf. In his
enthusiasm to research the market, he developed a
comprehensive seven-page questionnaire to be
completed by customers and distributed it among
surf supply stores.
a. How will the length of the questionnaire affect
the response rate?
b. What possible biases could arise from Mark’s
sampling method?
10. Families for the Future, a nonprofit organization
formed to promote a return to the family values of
the 1950s in the United States, is conducting a mail
survey to determine the level of domestic violence
and its causes. The organization plans to survey
families in urban areas in four cities across the
United States.
a. What are the possible sources of survey error in
this study?
b. How might the study be redesigned to eliminate
these errors?
End Notes
1. A. N. Oppenheim, Questionnaire Design and Attitude Measurement. New York: Basic Books, 1966.
2. George S. Day, “The Threats to Marketing
Research,” Journal of Marketing Research, 12, November 1975, pp. 462–467.
3. Frederick Wiseman and Mariane Schafer, “If Respondents Won’t Respond, Ask Nonrespondents
Why,” Marketing News, September 9, 1977, pp. 8–9;
Susan Kraft, “Who Slams the Door on Research?,”
American Demographics, September 1991, p. 14.
4. A. Ossip, “Likely Improvements in Data Collection
Methods—What Do They Mean for Day-to-Day
Research Management?,” Journal of Advertising
Research, Research Currents, October/November
1986, pp. RC9–RC12.
5. Arthur Saltzman, “Improving Response Rates in
Disk-by-Mail Surveys,” Marketing Research, 5, Summer 1993, pp. 32–39.
6. Charles F. Cannel, Lois Oksenberg, and Jean M.
Converse, “Striving for Response Accuracy:
Experiments in New Interviewing Techniques,”
Journal of Marketing Research, 14, August 1977,
pp. 306–315.
7. W. Cook, “Telescoping and Memory’s Other Tricks,”
Journal of Advertising Research, 87, February–March
1987, pp. 5–8.
8. James Julbert and Donald R. Lehmann, “Reducing
Error in Question and Scale Design: A Conceptual
Framework,” Decision Sciences, 6, January 1975,
pp. 166–173.
9. Bill Iuso, “Concept Testing: An Appropriate Approach,” Journal of Marketing Research, 12, May
1975, pp. 228–231.
10. Del I. Hawkins and Kenneth A. Coney, “Uninformed Response Error in Survey Research,” Journal
of Marketing Research, 13, August 1981, pp. 370–374.
228
11. William D. Wells, “The Influence of Yea Saying
Response Style,” Journal of Advertising Research, 1,
June 1963, pp. 8–18.
12. Donald P. Warwick and Charles A. Lininger, The
Sample Survey: Theory and Practice. New York:
McGraw-Hill, 1975, pp. 198, 203.
13. J. Freeman and E. W. Butler, “Some Sources of
Interviewer Variance in Surveys,” Public Opinion
Quarterly, Spring 1976, pp. 84–85.
14. Good discussions of what can be found in Warwick
and Lininger are in Robert Ferber, ed., Handbook of
Marketing Research, New York: McGraw-Hill, 1974,
pp. 2.124–2.132, 2.147–2.159.
15. This section is adapted from Floyd J. Fowler, Jr.,
Survey Research Methods. Newbury Park, CA: Sage,
1993.
16. T. Heberlein and R. Baumgartener, “Factors
Affecting Response Rates to Mailed Questionnaires: A Quantitative Analysis of the Published
Literature,” American Sociological Review, 1978,
pp. 447–462.
17. ARF Position Paper, “Phony or Misleading Polls,”
Journal of Advertising Research, Special Issue, 26,
January 1987, pp. RC3–RC8.
18. George S. Day, “The Threats to Marketing Research,” Journal of Marketing Research, 12, November
1975, pp. 462–467.
19. M. S. Frankel and S. Siang, “Ethical and Legal Aspects
of Human Subjects Research in Cyberspace. A Report
of a Workshop. June 10–11, 1999.” American Association for the Advancement of Science, November
1999, http://shr.aaas.org/projects/human_subjects/
cyberspace/report.pdf
20. J. Finn and M. Lavitt, “Computer Based Self-Help
Groups for Sexual Abuse Survivors,” Social Work
with Groups, 17, 1994, pp. 21–46.
CASE 9-1
Essex Markets
Essex Markets was a chain of supermarkets in a mediumsized California city. For six years it had provided its customers with unit pricing of grocery products. The unit
prices were provided in the form of shelf tags that showed
the price of the item and the unit price (the price per
ounce, for example). The program was costly. The tags had
to be prepared and updated. Further, because they tended
to become dislodged or moved, considerable effort was
required to make sure that they were current and in place.
A study was proposed to evaluate unit pricing.
Among the research questions in the study were the
following:
■ What percentage of shoppers used unit pricing?
■ With what frequency was unit pricing in use?
■ What types of shoppers used unit pricing?
■ For what product classes was it used most frequently?
■ Was it used to compare package sizes and brands or
to evaluate store-controlled labels?
It was determined that a five-page questionnaire
completed by around 1,000 shoppers would be needed.
The questionnaire could be completed in the store in
about 15 minutes, or the respondent could be asked to
complete it at home and mail it in.
Source: Prepared by Bruce McElroy and
David A. Aaker as a basis for class discussion.
■ What percentage of shoppers was aware of unit
pricing?
Questions for Discussion
Specify how the respondents should be approached in
the store. Write out the exact introductory remarks that
you would use. Should the interview be in the store, or
should the questionnaire be self-administered at home
and mailed in, or should some other strategy be
employed? If a self-administered questionnaire is used,
write an introduction to it. What could be done to
encourage a high response rate?
CASE 9-2
More Ethical Dilemmas in
Marketing Research
2.
The following scenarios are similar to those you saw in
Chapter 1, but bear directly on the rights of respondents in a research study. As before, your assignment
is to decide what action to take in each instance. Be prepared to justify your decision.
1.
Your company is supervising a study of restaurants
conducted for the Department of Corporate and Consumer Affairs. The data, which have already been
collected, included specific buying information and
prices paid. Respondent organizations have been
promised confidentiality. The ministry demands
that all responses be identified by business name.
Their rationale is that they plan to repeat the study
and wish to limit sampling error by returning to the
same respondents. Open bidding requires that the
government maintain control of the sample.
What action would you take?
3.
You are a project director on a study funded by a
somewhat unpopular federal policing agency. The
study is on marijuana use among young people in a
community and its relationship, if any, to crime. You
will be using a structured questionnaire to gather data
for the agency on marijuana use and criminal activities. You believe that if you reveal the name of the
funding agency and/or the actual purposes of the
study to respondents, you will seriously depress
response rates and thereby increase nonresponse bias.
What information would you disclose to
respondents?
You are employed by a market research company.
A manufacturer of female clothing has retained
your firm to conduct a study for them. The manufacturer wants you to know something about how
women choose clothing, such as blouses and
sweaters. The manufacturer wants to conduct group
interviews, supplemented by a session which
would be devoted to observing the women trying
on clothing, in order to discover which types of garments are chosen first, how thoroughly they touch
229
4.
and examine the clothing, and whether they look
for and read a label or price tag. The client suggests
that the observations be performed unobtrusively
by female observers at a local department store, via
a one-way mirror. One of your associates argues
that this would constitute an invasion of privacy.
What action would you take?
You are the market research director in a manufacturing company. The project director requests
permission to use ultraviolet ink in precoding
230
questionnaires on a mail survey. Although the
accompanying letter refers to a confidential survey,
the project director needs to be able to identify
respondents to permit adequate cross-tabulation of
the data and to save on postage costs if a second
mailing is required.
What action would you take?
Source: Provided with the permission of Professor
Charles Weinberg, University of British Columbia.
INFORMATION FROM
RESPONDENTS: SURVEY METHODS
10
Learning Objectives
■ Introduce the different kinds of survey methods.
■ Discuss the survey methods and describe the process involved using each method.
■ Enumerate the advantages and disadvantages of the survey method.
■ Discuss future trends in survey methods.
■ Elucidate the problems that the researcher faces in conducting international surveys.
The importance of reliable and timely data for marketing intelligence can be gauged if we realize how an uncertain future could make the plans of a corporation go haywire. It is thus
extremely important that the survey methods used be very reliable. There are as many survey
methods as there are different forms of communication technology. As the technology for communication progresses, the number of survey methods also proliferates. The web and recent
advances in PDA technology have introduced many new possibilities for conducting error-free
surveys. Hence, a researcher has to know the mechanics of each method clearly, and also how
it performs compared to the other methods. The choice among different survey methods is never
an easy one. The factors affecting the choice of a survey method were discussed in Chapter 9,
but without a thorough knowledge of the various methods, it will always be difficult to choose
among them. In this chapter, the three most prevalent methods of conducting surveys—personal
interviews, telephone interviews, and self-administered surveys—are discussed in detail. The latest trends in survey methods are also discussed. Conducting international surveys leads to a
number of new problems that are not encountered in domestic research. These problems and
possible solutions are also discussed briefly.
COLLECTING DATA
Recent studies have revealed that only about 10 to 15 percent of consumers contacted at home will
participate in a survey. American consumers are confronted by about 250 surveys (telephone, email
and self-administered surveys) a year. One approach to offset such an adverse reaction to surveys
is the relationship survey, a customer satisfaction survey that not only measures a vendor’s competency with the customer but also provides a closed-loop feedback process for sharing the results
with participants. The customer satisfaction survey can be turned into a relationship-building
process, and to pursue this opportunity, one must be prepared to deliver the survey results in a
detached, professional manner. Below are some guidelines for doing so:
1. Reviewing data—Carefully read the results of the survey, review all comments, and note any
common or consistent areas of dissatisfaction or satisfaction. Look at the information objectively.
2. Getting started—Create a list of the “good” and “bad” aspects of the survey results. Revise
your list as you see fit.
231
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
3. Setting the feedback objective—Before preparing the feedback vehicle, create a clear objective
statement.
4. Customer presentation—There are four basic ways the information can be shared with the participants: by newsletter, with a letter from the company, with a group presentation, and with
one-on-one presentations.
5. Sharing responsibility—Emphasize the joint-planning aspect as you get and keep the customers
involved. A joint effort will lead to better solutions.
6. Handling issues you cannot fix—Be honest with customers, and point out that if many customers have the same complaint, the problem will be immediately investigated as thoroughly
as possible.
7. Working the issue resolution with your account—Try to resolve any issues that have developed,
but remember to keep your promises realistic.1
Personal Interviewing
The different methods of conducting personal interviews can be classified based on the respondents to be contacted and on the means of contacting them. In this section the different types of
personal interview methods are discussed, as well as their advantages and limitations.
Process. The personal interviewing process is characterized by the interaction of four entities:
the researcher, the interviewer, the interviewee, and the interview environment. Each of the three
participants has certain basic characteristics, both inherent and acquired. Each also has general
research knowledge and experience, which vary a great deal among them. Collectively, these
characteristics influence the interviewing process and, ultimately, the interview itself. During a
personal interview, the interviewer and the interviewee interact and simultaneously influence
one another in an interview environment. The choice of an interview environment is made by
the researcher, depending on the type of data to be collected. A brief discussion of the various
personal interview methods, classified according to interview environment, follows.
At Home or Work Interviewing. The at home or work interview, in which consumers are interviewed in person in their homes or at their office, has traditionally been considered the best
survey method. This conclusion was based on a number of factors. First, the door-to-door interview is a personal, face-to-face interview with all the attendant advantages—feedback from the
respondent, the ability to explain complicated tasks, the ability to use special questionnaire techniques that require visual contact to speed up the interview or improve data quality, the ability to show the respondent product concepts and other stimuli for evaluation, and so on.
Second, the consumer is seen as being at ease in a familiar, comfortable, secure environment.
The door-to-door interview remains the only viable way to do long, in-depth, or detailed
interviews and certain in-home product tests. In addition, the door-to-door survey is the only
way currently available to obtain anything approaching a probability sample in a study that
involves showing concepts or other stimuli to consumers.
232
Executive Interviewing. The term executive interviewing is used by marketing researchers to refer
to the industrial equivalent of door-to-door interviewing. This type of survey involves interviewing business people at their offices concerning industrial products or services.
This type of interviewing is very expensive. First, individuals involved in the purchase decision for the product in question must be identified and located. Once a qualified person is
located, the next step is to get that person to agree to be interviewed and to set a time for the
interview.
Finally, an interviewer must go to the particular place at the appointed time. Long waits are
frequently encountered; cancellations are not uncommon. This type of survey requires the very
best interviewers, because frequently they must conduct interviews on topics about which they
know very little.
COLLECTING DATA
Mall Intercept Surveys. Shopping-center interviews are a popular solution when funds are limited and the respondent must see, feel, or taste something. Often, they are called shopping mall
intercept surveys, in recognition of the interviewing procedures. Interviewers, stationed at
entrances or selected locations in a mall, randomly approach respondents and either question
them at that location or invite them to be interviewed at a special facility in the mall. These facilities have equipment that is adaptable to virtually any demonstration requirement, including
interview rooms and booths, kitchens with food preparation areas, conference rooms for focus
groups, closed-circuit television and sound systems, monitoring systems with one-way mirrors,
and online video-screen interviewing terminals.
Since interviewers don’t travel and respondents are plentiful, survey costs are low. However,
shopping center users, who are not representative of the general population, visit the center with
different frequencies, and shop at different stores within the center.2 These problems can be minimized with the special sampling procedures described in Chapter 14. The number of people
who agree to be interviewed in the mall will increase with incentives.3
Mall intercept surveys, known for their low cost, have recently been plagued by cost/quality
issues. The desire to provide “better,” “faster,” and “cheaper” service has cost the research/opinion
industry money, quality, and efficiency of item and resources, and the satisfaction of the
customer.
In situations where significantly low incidence levels are expected or when the data collected
must meet special parameters for analysis or modeling, small bases (number of interviews in a
given location) are appropriate. However, key players are mistakenly choosing smaller bases (and
therefore, more locations on a single project), assuming that it will lead to faster completion of
the interviewing process and save money.
If purchasers of data collection services consider larger base sizes whenever possible, the
research process can be positively affected in the following ways:
■ Consistency in the number of quality interviewers working on the project.
■ Increased sample representatives at individual mall locations.
■ Decreased “real” costs—such as shipping, travel, phone, facsimile, and production of concepts,
prototypes, and videos—associated with more markets.
■ More efficient use of the available respondent pool at the mall.
■ Decreased briefing time necessary to train interviewers properly on each project.4
Purchase Intercept Technique.5 The purchase intercept technique (PIT) is different from but
related to the mall intercept approach. This technique combines both in-store observation and
in-store interviewing to assess shopping behavior and the reasons behind that behavior. Like a
mall intercept, the PIT involves intercepting consumers while they are in a shopping environment; however, the PIT is administered at the time of an observable, specific product selection,
as compared to consumers in a mall location. The researcher unobtrusively observes the customer make a purchase in a particular product category; then the researcher intercepts the customer for an interview as soon as the purchase has been made.
The major advantage of the PIT is that it aids buyer recall. Interviewing at the point of purchase minimizes the time lapse between the purchase and data collection, and can provide a
neutral set of memory cues for the respondent while the purchase is still salient. Apart from
difficulties in gaining access to stores, the principal disadvantage of the PIT is that it samples
only purchasers and not anyone else who might be influencing the decision on what to buy or
where to shop.
Omnibus Surveys. Omnibus surveys are regularly scheduled (weekly, monthly, or quarterly)
personal interview surveys with questions provided by a number of separate clients. The questionnaires, based on which the interviews are conducted, will contain sequences of questions on
different topics. Each sequence of questions is provided by one client, and the whole questionnaire is made up of such sequences of questions, on diverse topics, from different clients.
233
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
There are impressive advantages to the omnibus approach whenever only a limited number of personal interview questions is needed. The total costs are minimized, since the rates are
based on the number of questions asked and tabulated, and the cost of the survey is shared by
the clients. The results are available quickly, because all the steps are standardized and scheduled in advance. The regularity of the interview schedule and the assurance that the independent samples are matched make this a suitable base for continuous “tracking” studies and
before–after studies. Some omnibus operators offer split-run facilities, so that half of the sample receives one stimulus (one version of a question or concept) and the other, matched half,
gets another version. Also, by accumulating over several waves of interviews, it is possible to
conduct studies of low-incidence activities, such as the extent of salt-free diets and shortwave
transmitter ownership.
Advantages. An interviewer, face to face with a respondent, can do a great deal to arouse initial interest and thereby increase the rate of participation and continuing rapport. To reduce the
likelihood of a respondent refusing to finish the interview with an interviewer, it is also feasible to ask complex questions and enhance their meaning with pictorial and mechanical aids,
clarify misunderstandings, and probe for more complete answers. For these reasons, the personal interview usually is preferred when a large amount of information is required and the
questions are complex or involve tasks such as sorting cards into ordered piles or evaluating
visual cues such as pictures of product concepts or mock-ups of advertisements.
The personal interview questionnaire has a high degree of flexibility. For example, if the
answer to “Have you ever heard of (a community agency)?” is “yes,” the interviewer asks questions A and B, but if the answer is “no,” the interviewer asks about the next agency.
Generalizations about the accuracy of personal interview responses are hazardous. On one hand,
interviewer probes and clarifications maximize respondent understanding and yield complete
answers, especially to open-ended questions. Possibly offsetting these advantages are the problems of prestige seeking, social desirability, and courtesy biases discussed earlier. In relative
terms, it seems that for questions about neutral topics, all three methods are equally satisfactory,
whereas for embarrassing topics the personal interview is at a disadvantage.
Finally, there is an advantage to having a personal interview when an explicit or current
list of households or individuals is not available. The interviewer can be assigned to specific
census tracts, blocks, or residences as defined by census data. Once a residence is chosen, the
researcher can control who is interviewed and how much assistance is obtained from others
in the household.
234
Limitations. Personal interview studies are time consuming, administratively difficult, and
costly. The time requirements are understandable in light of the need to travel between interviews, set up appointments, and perhaps schedule return visits to complete interrupted interviews. Only 30 to 40 percent of an interviewer’s time on the job is devoted to interviewing
itself. One can use more interviewers to reduce the elapsed time, but then problems of quality control increase. Because of the time and administrative problems, the cost per completed
personal interview tends to be higher than it is for mail or telephone surveys. Direct comparisons of the costs of different methods are difficult, in part because of the wide variability in
the implementation of each method.
Table 10-1 provides some approximate indices of the direct cost of a completed interview, to
help compare data collection methods. In 1996, an index value of 1.0 corresponded to a cost of
$20.00. Thus, in 1996, one could expect a 40- to 60-minute personal interview on a national basis,
with one callback and 10 percent validation, to cost between $50 (2.5 ⫻ $20.00) and $70.00 (3.5 ⫻
$20.00). This cost assumes that the study is conducted by a commercial research supplier and a
general population is interviewed. If the sample to be interviewed consists of a particular segment, such as physicians, then a 40- to 60-minute interview would cost close to $200 per
interview.
Until recently, it was thought that the personal interview method was always the best way
to reduce nonresponse bias. Interviewers can track down hard-to-find respondents and minimize
COLLECTING DATA
TABLE 10-1
Comparative Indices of Direct Costs per Completed Interview
(Including Travel and Telephone Charges, Interviewer Compensation,
Training, and Direct Supervision Expenses)
Data Collection Method
Index of Costa
1. Email and Mail Survey (costs depend on return rate, incentives, and followup procedure).
0.3–0.8
2. Telephone interviews
a. Seven-minute interview with head of household in metropolitan area
b. Fifteen-minute interview with small segment of national population from
a central station
0.5–0.8
1.3–1.7
3. Personal interviews
a. Ten-minute personal interview in middle-class suburban area (1 call-back
and 10 percent validation)
b. Forty- to 60-minute interview of national probability sample (1 call-back
and 10 percent validation)
c. Executive (VIP) interviews
a
1.5–1.8
2.5–3.5
4.0–15.0⫹
In 1997, an index value of 1.0 corresponded to a cost of $20.00.
refusals by being physically present at the door. Unfortunately, the costs of the callbacks needed
to achieve high response rates are becoming excessive. This is especially a problem in inner-city
areas, which interviewers are reluctant to visit and may refuse to enter at night even when they
work in teams.
Screening questions are used in marketing research to find people with particular behavioral,
attitudinal, or demographic characteristics. Screening questions are usually asked first to minimize fieldwork costs or to enhance a study’s cost-effectiveness. Unfortunately, this could cause
self-selection bias, because the respondents receive clues as to what the researchers want to know
and they have the option to refuse to participate. “To the point” screening questions can offer a
signal as to who qualifies for the survey and who doesn’t.
Although a screener saves time and money by asking screening questions at the beginning
of the survey, as soon as self-selection bias occurs the survey is no longer random or representative. From the results of many studies involving funerals, careers, pets, and night clubs, it was
determined that it is wise to begin by working with a sample frame that will minimize the number of screening questions needed. When that is not practical, researchers should use more buffer
questions and write less obviously worded screeners to minimize self-selection bias. Although
this may increase costs, the findings will be more accurate and the research more valuable. Ignoring the problem will cause survey validity along with confidence in the work to deteriorate.6
Telephone Interviewing
The telephone interview gradually has become the dominant method for obtaining information
from large samples, as the cost and nonresponse problems of personal interviews have become
more acute. At the same time, many of the accepted limitations of telephone interviewing have
been shown to be of little significance for a large class of marketing problems.
Process. The telephone interviewing process generally is very similar to personal interviewing. Only
certain unique aspects of telephone interviewing, such as selecting the telephone numbers, the call
outcomes, the introduction, when to call, and call reports, are described below.
Selecting Telephone Numbers. There are three basic approaches to obtaining telephone numbers
when selecting study participants for telephone interviews. A researcher can use a prespecified list,
235
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
a directory, or a random dialing procedure. Prespecified lists—membership rosters, customer lists,
or lists purchased from commercial suppliers of telephone numbers—are sometimes used for
selected groups of people. This use, however, is not widespread in marketing research.
The traditional approach to obtaining numbers has been to use a directory, one provided by
either a telephone company or a commercial firm (for instance, the Polk crisscross directory).
However, a directory may be inadequate for obtaining a representative sample of consumers or
households. On the average, about 25 percent of the U.S. households that have telephones are
not included in telephone directories. People who are voluntarily not listed in a telephone directory tend to have characteristics somewhat different from those listed.
To overcome telephone directory nonrepresentativeness, many researchers now use randomdigit dialing when they interview consumers by telephone. In its most general form, complete
random-digit dialing is a nondirectory procedure for selecting all 10 (area code, prefix or
exchange, suffix) telephone number digits at random. Although this approach gives all households with telephones an approximately equal chance of being called, it has severe limitations.
It is costly to implement, both in dollars and time, since not all possible telephone numbers are
in service, and therefore many telephoning attempts are to nonexistent numbers. Additionally,
complete random-digit dialing does not discriminate between telephone numbers in which a
researcher is interested and those of no interest (numbers outside the geographic study area and
those of business or government). Further, all the prefixes in the United States are distributed
across the regular land lines (78 percent), cellular (10.5 percent), paging (4.6 percent), shared wireless (3.3 percent), and other (4 percent).7
A variation of the random-digit dialing procedure is systematic random-digit dialing (SRDD).
In SRDD, a researcher specifies those telephone area codes and exchanges, or prefixes, from
which numbers are to be selected. Thus, government, university, business, or exchanges not of
interest (outside the geographic study area) are avoided. The researcher determines a starting
number (seed point) plus a sampling interval—a constant number added systematically to the
starting number and subsequent numbers generated, to obtain the list of telephone numbers to
be called. For metropolitan areas, it is necessary to generate approximately four times as many
telephone numbers as completed interviews desired, because of not-in-service numbers and the
like. An illustration of the SRDD process is described in Marketing Research in Action 10-1.
SRDD has several advantages. Because there is a random starting point, each telephone number has an equal chance of being called. Second, since telephone exchanges tend to cover specific geographic areas, a spatial focus is possible. Third, if the same number of telephone calls is
attempted from each exchange studied, the resulting sample tends to have the same geographic
dispersion as the original population. Finally, SRDD can be incorporated into a standard computer program.
Another version of random-digit dialing is plus-one dialing, a directory- assisted, randomdigit-dialing telephone number selection procedure. Plus-one dialing consists of selecting a random
sample of telephone numbers from one or more telephone directories, then adding the constant “1”
to the last four digits of each number selected. This procedure increases the chances of an existing
telephone number being obtained and also allows unlisted numbers to be included in the sample.
MARKETING RESEARCH IN ACTION 10-1
An Illustration of the SRDD Process
Suppose an interviewer wants to poll a thousand (n ⫽ 1,000) respondents on the eve of the presidential
election in a particular area (with the area code prefix as 743). There are a total of ten thousand (k ⫽
10,000) numbers with the prefix 743, that is, 743-0000 to 743-9999. The first step in the SRDD process is
to compute the sampling interval (I) given by k兾n, which in this case is equal to 10. The interviewer then
randomly chooses a telephone number in the interval 743-0000 to 743-0010. Once a number is chosen
(say, 743-0005), then, to generate additional numbers, the value of I is added to each of the previously
selected numbers. In other words, the telephone numbers to call would be 743-0005, 743-(0005 ⫹ I),
743-(0005 ⫹ 2I), . . . ., 743-[0005 ⫹ (n ⫺ 1)I].
236
COLLECTING DATA
TABLE 10-2
Call Outcomes and Recommendations to Deal With Them
Call Outcome
Recommendation
The telephone is not in service.
Eliminate the number from further consideration.
The number dialed is busy.
Call the number again later, because the characteristics of
the people whose lines are busy will be different from those
whose lines are not.
No one answers the call.
Call the number back later, because the characteristics of
the people who are not at home will be different from those
who are at home.
The number called is a fax
number.
Send a fax to the respondent requesting his or her time to
conduct the interview, and get his or her telephone number.
An answering machine
comes on.
Leave a message on the answering machine saying who you
are and the purpose of your call. Call the number again after
some time.
The call is answered by someone other than the respondent.
Find out when the respondent will be available and call back
at that time.
The person contacted is not in
the sampling frame.
Eliminate the number from further consideration.
The call is answered by the
person to be contacted.
Conduct the interview.
Call Outcomes. Once the telephone numbers have been selected, a call is made. Once a call has
been attempted, eight possible outcomes can occur. The possible outcomes of a call attempt are
as follows: The telephone number is not in service, the number dialed is busy, no one answers
the phone, the number called is a fax number, an answering machine responds, the call is
answered by someone other than the person to be contacted, the call results in contacting a person outside the sampling frame (the telephone line was given to someone else when the respondent relocated), and, finally, the call is answered by the actual respondent. Table 10-2 discusses
the various call outcomes and the method to handle each outcome.
The Introduction. One of the most important aspects of telephone interviewing, and the key
to a successfully completed interview, is the introduction. For the interview to be completed
successfully, the interviewer must gain immediate rapport with potential study participants.
Gaining rapport requires a pleasant telephone voice (being male or female does not seem to
matter) and a good introduction. It is important that the introduction introduce the topic of
study and be brief. An overly long introduction tends to decrease cooperation and elicit refusals
to participate.
When to Call. To efficiently obtain a representative sample of study participants, telephone interviews should be attempted at times when prospective interviewees will most likely be available.
For consumer interviews, telephone interviews should probably be attempted between 6 P.M. and
9 P.M. on weekdays, and 10 A.M. to 8 P.M. on weekends. Calling before 6 P.M. on weekdays
decreases the chances of reaching working individuals, and calling after 9 P.M. incurs the wrath
of those who are early to bed. On the other hand, the best time to reach homemakers or contact
individuals at work is between 9 A.M. and 4:30 P.M.
Call Reports. A call report is a form that has telephone numbers to be called and columns for
interviewers to document their telephoning attempts—what day and time the call was made,
the outcome, the length of the call, and so forth. Call reports provide records of calling experiences and are useful for managing data collection.
237
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
Advantages. Telephone interviews may be conducted either from a central location, at prescribed
hours under close supervision, or from the interviewer’s home, unsupervised and at his or her
own hours. The former is preferred because problems can be isolated quickly and greater uniformity is possible. Supervisors can double-record interviews by listening on an extension and
can gradually weed out incompetent interviewers.
Regardless of how the telephone interviews are conducted, the obvious advantages are the
same: (1) More interviews can be conducted in a given time period, because no time is lost in
traveling and locating respondents; (2) more hours of the day are productive, especially the
evening hours when working women and singles are likely to be at home and apartment doors
are locked; and (3) repeated callbacks at different times of the day can be made at very low cost.
The key to the low costs for the latter surveys is the wide-area telephone service (WATS), which provides unlimited calls to a given zone in Canada or the United States for a fixed monthly charge.
Overall, the telephone method dominates the personal interview with respect to speed,
absence of administrative problems, and cost per completed interview. As we saw in Table 10-1,
the costs of a telephone survey seldom will exceed two-thirds of the comparable costs of a personal interview.8 Costs can be reduced further with omnibus surveys.
For better or worse, the telephone is an “irresistible intruder.” A ringing telephone literally compels us to answer. Long-distance calling brings a further dimension of urgency and
importance to reaching the desired respondent. This tends to counteract the fact that it is
easier for a person to terminate midway through a telephone interview. The telephone is a particularly effective method for gaining access to hard-to-reach people such as busy executives.
The receptionist who thwarts the personal interviewer will readily connect a telephone interview request. Thus, Payne found that 94 percent of the interviews attempted with responsible
persons in the 600 largest companies were completed successfully.9 The intrusiveness of the
telephone, plus the ease of making callbacks, means that there should be less sample bias due
to nonresponse.
For most topics there is likely to be little difference in the accuracy of responses between telephone and personal interviews. The Survey Research Center (University of Michigan) found that
similar aggregate results were obtained by the two methods in their quarterly interview of consumer
intentions. However, less differentiated responses were obtained over the telephone.10
During the telephone interview, the respondent’s only impression of the interviewer is that
conveyed by the voice. The lack of rapport is offset to a degree by lessened interviewer bias and
greater anonymity of the situation. However, the research on whether the respondent will reply
with greater candor to personal questions (such as alcohol consumption) is mixed.
238
Limitations. Relatively few of the problems with the telephone method are completely insurmountable. The most obvious problem is the inability to employ visual aids or complex tasks. For
example, it does not appear feasible to ask respondents to retain in their minds the names of nine
department stores and then ask them to choose one store. There have been solutions to this problem, including separate mention of individual stores and asking the respondent to treat the telephone push-buttons as a 10-point rating scale (from “1” for like it very much to “0” for don’t like it
at all). A related problem with the telephone is that the interviewer must rely solely on verbal cues
to judge the reaction and understanding of respondents.
There is some controversy about the amount of information that can be collected in a telephone interview. Most telephone interviews are kept as short as 5 to 10 minutes, because of the
belief that a bored or hurried respondent will be likely to hang up the phone. However, there is
a tendency for respondents to underestimate the length of time spent on a telephone call. Therefore, 20- to 30-minute interviews are increasingly frequent and successful, but only with interesting topics and capable interviewers.
A further limitation of telephone interviewing is the potential for sample bias, which is a consequence of some people being without phones, having unlisted phones, and telephone directories
being unable to keep up with a mobile population. A subsequent chapter (chapter 14) on sampling
will discuss some of the solutions to these sampling problems, as well as the use of callbacks to
reduce the frequency of not-at-homes.
COLLECTING DATA
Finally, a recent National Health Interview Survey explored the efficacy of random digit dialing.11 During the first half of 2006, nearly 12.5 percent of American households did not have a landline phone. Among these people, 84 percent had at least one wireless phone; and for some other
segments, wireless is the primary mode of communications. For instance, in targeting consumers
aged between 18 and 24, nearly 25 percent have only a wireless phone. This means that a random
digit dialing program would not be able to reach these consumers. Similarly, 44 percent of the
swinging singles consumer segment uses cell phones. These new findings have wide implications
on targeting consumers.
Self-Administered Surveys
The third major survey method is the self-administered survey. This type of survey can be
emailed, mailed, faxed or simply handed to the respondent. In the self-administered interview
method, no interviewer is involved. Even though this reduces the cost of the interview process,
this technique has one major disadvantage: There is no one present to explain things to the
respondent and clarify responses to open-ended questions. This results in the answers to most
of the open-ended questions being totally useless. Some have argued, however, that the absence
of an interviewer results in the elimination of interviewer bias.
Self-administered interviews are often used in malls or other central locations where the
researcher has access to a captive audience. Airlines frequently use this technique to get information about their services; the questionnaires are administered in flight. Many hotels, restaurants,
and other service businesses provide brief questionnaires to patrons, to find out how they feel about
the quality of service provided.
The Process. Using a self-administered survey consists of identifying and locating potential study
participants, deciding on the best way to get the questionnaire to those participants, and waiting
for completed questionnaires to be returned. Substantively, the process is a series of distinct and
often difficult decisions regarding the identification of study participants and the interview package. For example, for a mail survey you would need an outgoing envelope, cover letter, questionnaire, return envelope, and the incentives, if any, to be used.
Unlike personal and telephone interviews, self-administered interviews require at least
broad identification of the individuals to be sampled before data collection begins. Therefore,
an initial task is to obtain a valid mail or email address, fax number or location to distribute
the questionnaire to the study participants. This information can be obtained from customer
lists, association or organization membership rosters, telephone directories, publication subscription lists, or commercial “list houses.” Regardless of its source, the information must be current and relate closely to the group being studied. As might be expected, obtaining a useful
list is especially difficult when a list that is representative of the general population is desired.
Decisions must also be made about the various elements to be contained in a self-administered survey. Although many of these decisions involve relatively mechanical elements, each
decision influences both the response rate and the response quality. Some decisions that are to
be considered are
Method of addressing
Cover letter
Questionnaire length, content, layout, color, and format
Method of notification; should there be a follow-up?
Incentive to be given
One decision area is method of notification. In both preliminary and follow-up notification, a
researcher communicates with potential self-administered interview respondents more than once.
Preliminary notification may also be used to screen, or qualify, individuals for study inclusion. This
communication may be by postcard, letter, email or telephone. Follow-up communication methods
include sending potential study participants postcards or letters, or telephoning them a few days
after they receive the questionnaire to remind them to complete and return it as soon as possible.
239
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
Advantages. The most likely reason for choosing a mail survey is cost, but there are other
reasons also, as illustrated by the decision of the Census Bureau to switch from a personal
interview to a census by mail to gain “better results, including a shortening of the period for
collecting the data and more reliable answers supplied directly by respondents instead of
through a more-or-less inhibiting intermediary, the enumerator.”12 This approach worked well
for the 2010 census, when forms were mailed to over 90 percent of American households, of
which 90 percent were completed voluntarily. More recent tests, to larger segments of the population, have not been so encouraging.
There is consistent evidence that self-administered surveys yield more accurate results—
among those completing the survey. Because the questionnaire is answered at the respondent’s
discretion, the replies are likely to be more thoughtful and others can be consulted for necessary
information. Self-administered surveys generally are superior when sensitive or potentially
embarrassing topics, such as sexual behavior and finances, are covered (so long as the respondent
is convinced that the answers will be kept in confidence). For example, a study of Boston residents, which compared the three basic methods of data collection, found that each of them gave
equivalent results for neutral topics. On sensitive topics, however, where socially undesirable
responses were possible, there were large differences. On the question of legalizing abortion,
89 percent of mail survey respondents were in favor, compared to 62 and 70 percent of telephone
and personal interview respondents, respectively.13
Limitations. The absence of an interviewer means that a large number of variables are controlled
inadequately, including the following:
■ The identity of the respondent (Was it the addressee who answered, or someone else?)
■ Whom the respondent consults for help in answering questions
■ The speed of the response (The usual time lag before receipt of a questionnaire delays the
study and makes the responses vulnerable to external events taking place during the study.)14
■ The order in which the questions are exposed and answered (The respondent can look ahead
to see where the questions are leading, so it is not possible to funnel questions from the general to the specific, for example.)
■ Respondent understanding of the questions (There is no opportunity to seek clarification of
confusing questions or terms, so many respondents return their questionnaire partially
completed.)
240
One consequence of these problems is that long questionnaires with complicated questions
cannot be used without diminishing the response rate. As a rule of thumb, six to eight pages is
the upper limit on topics of average interest to respondents.15
Self-administered surveys are limited to situations where a list (mailing addresses, email
addresses, fax numbers, etc.) is available and the cost of the list is not prohibitive. Unfortunately,
there are a number of possible flaws in all such lists: obsolescence, omissions, duplications, and
so forth. This makes it difficult to find the ideal list, which consists entirely of the type of person to be contacted, and also represents all of those who exist. Great care must be taken at this
stage to ensure that the study objectives can be achieved without excessive compromise.
If the benefit of self-administered surveys is cost, then the bane is response rates. The response
rate gives an indication of the number of questionnaires that have been returned. The Council
of American Survey Research Organizations defines response rates as the ratio of the number of
completed interviews with responding units to number of eligible responding units in the sample.16 Implementing the definition may be simple or complex, depending on the methods used
to select the sample.17 The problem is not that acceptable response rates cannot be achieved, but
rather that the rate is hard to forecast and there is substantial risk that an acceptable rate may
not be achieved. Marketing Research in Action 10-2 illustrates how refusal rates are calculated.
Many factors combine to influence the response rate, including (1) the perceived amount of
work required, which in turn depends on the length of the questionnaire and the apparent ease
with which it can be completed; (2) the intrinsic interest of the topic; (3) the characteristics of the
COLLECTING DATA
MARKETING RESEARCH IN ACTION 10-2
Respondent Cooperation: Audit Update Yields
Ways to Improve Results
In early 1997, CMOR undertook a project to update refusal rate information. Forty CASRO and CMOR
member firms were recruited to keep track of all their phone research for April 1997. This yielded 385 surveys with a total base of 243,597 interviews for which sample disposition data and some basic facts were
provided about the subject matter and purpose of the survey, length of interview, client identification, type
of samples, and use of incentives. Refusal rate was calculated using the industry-accepted definition as
follows:
Total Refused
Refusal Rate
⫽
Total Contacts
Total Refused ⫽ Initial Refusals ⫹ Breakoffs ⫹ Qualified Refusals
Total Contacted ⫽ Total Refused ⫹ Eligible Respondents Not Available ⫹ Language Barrier
⫹ Completed Interviews
The result was that keeping the interview length short improves cooperation:
Interview
Refusal Rate
5 minutes or less
32%
10 minutes
45%
15 minutes
46%
20 minutes
56%
Over 20 minutes
57%
Disclosing the length of the interview in the “intro” seems to improve cooperation regardless of the
length of the interview (the base sizes for disclosed are small):
Refusal Rate
Interview
Length Disclosed
Length Not Disclosed
5 minutes or less
18%
38%
6–10 minutes
35%
46%
11–15 minutes
41%
47%
16 minutes or more
42%
55%
Honest disclosure of facts during interview introduction has a positive effect on cooperation:
Refusal Rate
Interview Length Disclosed
34%
Length Not Disclosed
52%
Client Identified
29%
Client Not Identified
54%
Subject Matter Disclosed
43%
Subject Not Disclosed
61%
However, providing reassurance in the interview introduction of whether the survey is confidential or
a sales pitch did not appear to improve cooperation:
Refusal Rate
Not a Sales Attempt
51%
No Mention of Sales
43%
Responses Confidential
48%
Confidentiality Not Mentioned
46%
SOURCE: Pam Humbaugh, Opinion—The News Magazine of CMOR, Vol. 1, No. 1, Winter 1998.
241
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
MARKETING RESEARCH IN ACTION 10-3
Ethics in Surveys
Millions of Americans find in the welter of junk mail they receive items labeled “survey,” usually in
oversized envelopes, sent by political or other advocacy groups. They usually include a brief questionnaire, perhaps half a dozen questions worded to encourage “yes” answers if the group is for something
or “no” answers if it is against. Most consumers can detect the manipulative nature of these surveys and
do not respond. Consumer nonresponse represents a threat to legitimate marketing research.
Though many marketing groups have spoken out against the unethical behavior of the causes and
candidates, public relations (PR) is most effective in stopping the distribution of advocacy surveys. After
authorizing a sound PR budget, mass media should be used to send the message about how surveys are
fooling the public by misrepresenting responses and asking for contributions under false pretenses.
Clients, the users of survey research information, should be made aware that discouraging the phonysurvey practice is crucial to them. These “surveys” are likely to undermine the reliability of the marketing research data they depend on by further discouraging consumer response. Their support will be
useful for the campaign against advocacy surveys.
SOURCE: Thomas T. Semon, “Advocacy Surveys Threaten Marketing Research,” Marketing News, April 24, 1995, p. 9.
sample; (4) the credibility of the sponsoring organization; (5) the level of induced motivation;
and (6) increased unethical use of marketing research to sell products. (See Marketing Research
in Action 10-3 for an example.) A poorly planned mail survey on a low-interest topic may achieve
only a 10 to 15 percent response rate. Under the right circumstances, 90 percent response rates
are also possible.
Average response
to question
Coping with Nonresponse to Self-administered Surveys. Nonresponse is a problem, because
those who respond are likely to differ substantially from those who do not respond. The best way
to protect against this bias is to improve the response rate. The most consistently effective methods for achieving high response rates involve some combination of monetary incentives and
follow-ups or reminders.18 The inclusion of a dollar bill in the mailing, which is the usual reward,
has been found to improve response rates by increments of 18 to 27 percent when compared to
returns when no incentive is used. Comparable improvements have been obtained from the single
or multiple follow-up letter. Although each follow-up brings additional responses, the optimum
number seems to be two. It is a moot point whether it is worthwhile to include another questionnaire in the follow-up letter. Other techniques for improving response rates, such as having a persuasive cover letter, appear to have a lesser but still worthwhile effect. Surprisingly, there is no
clear evidence that personalization of the cover letter or email, promises of anonymity, color, and
methods of reproduction make much difference.
Another approach to the nonresponse problem is to determine the extent and direction of
bias by studying the differences between those who respond and those who do not. This
sometimes can be done by taking a subsample of the nonrespondents and using a variety of
methods to get a high response rate from this group. Of course, when the questionnaire is
anonymous or time is short, this cannot be done. In this situation it may be possible to compare the results of the survey with “known” values for the population, using such variables
as age and income. Alternatively, one can use extrapolation methods, which assume that those
242
Wave 2
Last
respondent
Wave 1
First
wave
Second Nonrespondents
wave
Cumulative response rate
Project
response of
average
nonresponder
100%
COLLECTING DATA
who respond less readily are more like nonrespondents.19 “Less readily” can be defined as
either being slower than average in answering a simple mailing, or responding to the extra
prodding of a follow-up mailing. With data from two waves of mailings, a trend can be established in the pattern of answers; nonrespondents can be assumed to be like either the last
respondent to the second wave or like a projected respondent at the midpoint of the nonresponse group. This is represented graphically in the response rate chart.
Panels. A panel is a representative national sample of people who have agreed to participate
in a limited number of surveys each year. Some panels mail surveys to participants, others
have web-based portals that participants use to answer survey questions. A number of these
panels are operated by firms such as Market Facts, Home Testing Institute, and National Family Opinion, Inc. The latter firm, for example, offers a number of panels that contain 130,000
people. The major advantage is the high response rate, which averages 75 to 85 percent.
A typical mail panel is the Conference Board Survey of Consumer Confidence. Each quarter, a
large sample, drawn from the National Family Opinion Panel, is sent a questionnaire in the card
format shown in Figure 10-1. This card is sent with cards from as many as 10 other studies, which
spreads the costs of the survey across a number of companies. Since the questions and the sampling
procedures are the same every quarter, the Conference Board has a standard measuring stick for
tracking fluctuations in consumer attitudes and buying intentions.
Panels are recruited to match the general population with respect to geographic location, city
size, age of homemaker, family income, and so on. Hence, it is possible to draw special samples of
particular occupation groups (such as lawyers), age categories (such as teenagers), and geographic
areas. Large samples can be obtained quickly in test-market areas, for example. Inevitably, those
Answer This Side First
Answer Other Side First
1. How would you rate the present general business
conditions in your area?
䊐 GOOD
䊐 NORMAL
䊐 BAD
a. SIX MONTHS from now do you think they will be:
䊐 BETTER?
䊐 SAME?
䊐 WORSE?
6. Please check which, if any, of the items you plan to buy
in the next SIX MONTHS, and which brand you are
most likely to choose:
BRAND PREFERENCE
Refrigerator ....................01䊐 ________________________
Washing Machine ..........02䊐 ________________________
Black/White TV ..............03䊐 ________________________
Color TV ........................04䊐 ________________________
Vacuum Cleaner ..............05䊐 ________________________
Range ............................06䊐 ________________________
Clothes Dryer ..................07䊐 ________________________
Air Conditioner ..............08䊐 ________________________
Dishwasher......................09䊐 ________________________
Microwave Oven ............10䊐 ________________________
Sewing Machine..............11䊐 ________________________
Carpet (over 4⬘ ⫻ 6⬘) ....12䊐 ________________________
NONE OF THESE 䊐
2. What would you say about available jobs in your area
right now?
䊐 PLENTY 䊐 NOT SO MANY 䊐 HARD TO GET
a. SIX MONTHS from now do you think there will be:
䊐 MORE?
䊐 SAME?
䊐 FEWER?
3. How would you guess your total family income to be
SIX MONTHS from now?
䊐 HIGHER
䊐 SAME
䊐 LOWER
4. Does anyone in your household plan to buy a house in
the next SIX MONTHS?
䊐 YES
䊐 NO
䊐 MAYBE
a. If YES: 䊐 NEW?
䊐 LIVED IN? 䊐 DON’T
KNOW?
5. Does anyone in your household plan to buy a car in
the next SIX MONTHS?
䊐 YES
䊐 NO
䊐 MAYBE
a. If YES: 䊐 NEW?
䊐 USED?
䊐 DON’T
KNOW?
MAKE? _______________________________ 䊐 DON’T
KNOW?
CONTINUE S
7. Do you plan to take a vacation away from home
between NOW and the next SIX MONTHS?
䊐 YES
䊐 NO
䊐 UNDECIDED
a. Where will you spend most of your time while on
vacation?
䊐 HOME䊐 OTHER
䊐 FOREIGN
STATE
STATE(S)
COUNTRY
b. How will you mainly travel?
AIRPLANE
CAR
TRAIN
BOAT
BUS
¬¬ 䊐
䊐
䊐
䊐
䊐
T
FIGURE 10-1
Conference Board mail panel survey of consumer confidence.
243
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
people who agree to serve on such panels will be different from the rest of the population—perhaps
because they are more interested in such research, have more time available, and so forth. Little is
known about the impact of such differences on questionnaire responses.
Web Surveys20
Web-based marketing research studies fundamentally and permanently are changing the role
marketing research will play in future businesses. Web-based surveys have many advantages
over traditional methods: they can be of high quality, fast, and inexpensive.
There are no interviewers involved, thereby eliminating interviewer error and bias. Webbased surveys give researchers much more control over data quality. Logic checks can be built
into the survey so that contradictory or nonsensical answers are not allowed, eliminating the
need for data cleaning and editing. With a Web-based survey, the questionnaire is posted on a
secure website where clients can instantaneously view results as soon as any respondent has
completed the interview. Web-based marketing can also reduce data collection costs to zero; this
could be done by posting invitation banners on company websites that receive high traffic from
target customers. These website visitors would click on the banner and immediately be sent to
a secure website to conduct a brief interview. Afterward, respondents are automatically returned
to their point of departure.
On the other hand, the validity of online surveys depends on the sample selection, survey
design, response tendencies, and technology challenges.
In e-research, the choice of sampling units is in the form of e-mail addresses, electronic subscription groups, and heavily visited websites. Unfortunately, people change their e-mail
addresses very often, thus creating problems in obtaining well-defined samples. Another problem
is the poor translation of well-designed surveys into electronic versions. This results in lower
response rates for the surveys, which in turn lower their statistical value and validity.
Therefore, Web-based surveys need continuous monitoring to determine their direction,
appropriate use, and long-term effectiveness. Research is being transformed by the online
medium; however, and increasing numbers of people are opting for fast and efficient Web-based
surveys.21 Web-based marketing research will greatly increase the amount of customer feedback
on which managers base critical business decisions.
Combinations of Survey Methods
Since each of the basic methods of data collection has different strengths and weaknesses, it is
sometimes desirable to combine them and retain the best features of each while minimizing the
limitations. Some of the feasible combinations (or sequences) are illustrated below:
1. Telephone ——————›
appointment
Personal interview ———–——›
2. Telephone ——————›
request
Mail survey ———–—————›
for permission
to mail a questionnaire
Telephone pre-interview
(as part of a panel study
or verification)
Telephone follow-up
Self-administered ————–––—›
questionnaire
delivered by
“interviewer”
Fax survey –——————–——›
telephone (optional)
3. Personal ———————›
interview
4. Mail survey –—————›
5. Telephone or —————›
personal
244
6. Telephone request –—–—›
for permission
to fax questionnaire
Leave behind a
self-administered
questionnaire to be
mailed later
Telephone follow-up
(optional)
To be either
picked up or
mailed in
Follow-up
COLLECTING DATA
With the exception of the reinterview panel design of sequence 3, each of the above combinations has proven very effective in increasing the response rate. Indeed, sequence 1 is virtually
mandatory for personal interviews with executives, as we noted in the previous chapter. The
virtues of the other sequences are not quite so obvious. Marketing Research in Action 10-4
describes a study that included a combination of survey methods, namely a mail survey with a
telephone follow-up.
The telephone prenotification approach is essentially a phone call to ask permission to mail or
fax a questionnaire. The key is the telephone presentation, which must not only gain agreement
to participate but also make sure the prospective respondent is serious about cooperating. In one
study in which 300 households were phoned, 264 households (88 percent) agreed to the mailing
and 180 (60 percent) returned usable questionnaires.22 If the return rate is not acceptable, a
follow-up phone call can be made.
The lockbox approach is designed to circumvent the screens that receptionists and
secretaries set around busy executives. The mail is used to deliver a small locked metal box
containing a questionnaire and other interviewing materials such as flashcards, exhibits, and
pictures. A covering letter, attached to the box, explains the purpose of the survey and tells
the prospective respondent that an interviewer will conduct a telephone interview in a few
days. The letter also tells the respondent that the box is a gift but that the combination to the
lock will not be provided until the time of the interview. Ostensibly, this is so the respondent
will not be biased by seeing the interview materials in advance. However, it is clear that the
value of the gift depends on participation in the survey. Also, the locked box stimulates
respondent curiosity. As a result, the originator of this technique has been able to obtain
response rates of 50 to 70 percent, even with notoriously difficult respondents such as
attorneys in private practice, who sell time for a living and are loath to participate in “free
interviews.”23
The drop-off approach is an illustration of sequence 5, which is particularly well suited to
studies within compact geographic areas. For public transit studies, the questionnaire can be hand
delivered to sampling points, such as every 25th house within areas that have access to a transit
line. It also can be used for subjects living in designated political precincts or within a given radius
of a specific retail outlet or other service. The major advantages are that: (1) only lightly trained
interviewers are required to gain the cooperation of the respondents, deliver the questionnaires,
and arrange a return visit; (2) response rates are high, generally between 70 and 80 percent, in part
because of the initial commitment to cooperate, coupled with the realization that the person who
dropped the survey will be returning to pick up the completed questionnaire; (3) several
questionnaires can be left in each household, if all adults are part of the sample; (4) lengthy
questionnaires can be used without affecting the response rate; and (5) it is a very cost-effective
method.24
MARKETING RESEARCH IN ACTION 10-4
Use of Mail Survey with Telephone Follow-up
A study was conducted to determine the effect on response rates when mail and telephone survey methods were used. The survey also included a prepaid $2.00 cash incentive. The survey drew samples from
Medicaid enrollees, who were stratified by race and ethnicity.
The sampling was performed in two stages. While the first stage consisted of a simple random sample (SRS) of Medicaid enrollees, the second stage consisted of randomly drawn samples from American
Indian, African-American, Latino, Hmong, and Somali enrollees. A total of 8,412 enrollees were assigned
randomly to receive a mail survey with no incentive or a $2.00 bill.
The response rate within the SRS following the mailing portion of the survey was found to be 54 percent
in the incentive group and 45 percent in the nonincentive group. After telephone follow-ups, the incentive SRS
response rate increased to 69 percent, and the nonincentive response rate increased to 64 percent. It was seen
that the inclusion of the $2.00 incentive had similar effects on response rates.
SOURCE: Timothy Beebe, Michael Davern, Donna McAlpine, Kathleen Call, and Todd Rockwood. “Increasing Response Rates in a Survey of Medicaid
Enrollees: The Effect of a Prepaid Monetary Incentive and Mixed Modes (Mail and Telephone),” Medical Care, 43(4); April 2005, pp. 411–414.
245
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
Table 10-3 presents a comprehensive set of advantages and disadvantages of the various
methods discussed above.
Prepaid phone cards are now being pitched as a way to get consumers to participate in
surveys for marketing research. MHA Communications of Illinois has designed a system for
gathering consumer information through an interactive voice response (IVR) system linked to
a phone card.
Certain consumers are given prepaid phone cards worth about 10 minutes by a client company. Once they call the 800 number, the callers are offered additional free phone time in
exchange for answering a few survey questions. One advantage of IVR phone card research is
flexibility and convenience for its customers; they can call whenever they like. IVR is also flexible enough to allow companies to alter survey questions and substitute new ones. But for
consistency, only data from the new questions should be measured, and data from ineffective
questions dropped.25
Trends in Survey Research
Surveys will continue to dominate as a method of data collection for marketing research, at least
for the next few decades. Rapid advancement in technology, however, is changing the very nature
of data collection and survey methods. Computers and interactive technology are revolutionizing the way surveys are conducted. This section will look into some survey methods that have
become popular in the last few years, and also the future of survey research.
Computer-Assisted Telephone Interviewing (CATI). Computers are being used increasingly
to control the administration and sequence of questions asked by an interviewer seated at a
terminal. This use of computers provides researchers with a way to prevent many interviewer
errors, such as choosing the wrong respondent in a household, failing to ask a question that
should be asked, or asking a series of questions that is not appropriate for a particular respondent. If the respondent selection procedure or the skip patterns in a questionnaire are complicated, these interviewer errors are more likely to happen. When control of the question
sequence is turned over to the computer, the interviewer theoretically is free to concentrate
fully on reading the questions, recording the answers, and establishing rapport with the
respondent.
There are some limitations to computer-controlled telephone interviewing. It is generally
more expensive to use the computer than to administer the traditional paper-and-pencil questionnaire. For that reason, use of a computer-controlled system is recommended when a large
number of surveys must be done or when the questionnaire will be used many times in a tracking study.
A second limitation relates to the problems involved in using a mechanical system and to
human error in its programming and operation. A program must be written and carefully
debugged for each questionnaire. The computer system must be able to handle the demands
of a large number of interviewers. At the worst, several days of interviewing may pass before
someone notices that a mistake has been made and the questionnaire program is incorrectly
skipping over a crucial series of questions. Valuable interviewing time and money may be
lost if the computer system becomes overloaded and “crashes.”
At its best, computer-controlled telephone interviewing can produce faster, more complete,
data to the researcher.
Computer Interactive Interviewing. In computer interactive interviewing the respondent interacts directly with the computer. The respondent is asked to sit before the computer and to
answer the questions as they appear on the screen. This type of interview is being widely used
in malls. Customers are intercepted and qualified in the mall and then brought to a test facility in the mall. They are seated at a computer terminal, given some instruction on what to do,
246
COLLECTING DATA
TABLE 10-3
Advantages and Disadvantages of Survey Methods
Type of
Survey
Method
Personal
interviewing
Advantages
■
■
■
■
■
■
Telephone
interviewing
■
■
■
■
■
■
■
Selfadministration
(mail, email,
fax, drop-off,
etc.)
■
■
■
■
■
■
■
Disadvantages
There are sample designs that can be implemented best by personal interview (for
example, area probability samples).
Personal interview procedures are probably
the most effective way of enlisting
cooperation.
Advantages of interview questions—probing
for adequate answers, accurately following
complex instructions or sequences—are
realized.
Multimethod data collection, including observation, visual cues, and self-administered
sections, are feasible.
Rapport and confidence building are possible
(including any written reassurances that may
be needed for reporting very sensitive
material).
Probably longer interviews can be done in
person.
■
Lower costs than personal interviews.
Random-digit-dialing (RDD) sampling of
general population.
Better access to certain populations, especially as compared to personal interviews.
Shorter data collection periods.
The advantages of interviewer administration (in contrast to mail surveys).
Interviewer staffing and management easier
than personal interviews—smaller staff
needed, not necessary to be near sample,
supervision and quality control potentially
better.
Likely better response rate from a list
sample than from mail.
■
Ease of presenting questions requiring
visual aids (in contrast to telephone
interviews).
Asking questions with long or complex
response categories is facilitated.
Asking batteries of similar questions is
possible. The respondent does not have to
share answers with an interviewer.
Relatively low cost.
Can be accomplished with minimal staff
and facilities.
Provides access to widely dispersed
samples and samples that for other reasons
are difficult to reach by telephone or in
person.
Respondents have time to give thoughtful
answers, look up records, or consult with
others.
■
■
■
■
■
■
■
■
■
■
■
■
■
It is likely to be more costly than
alternatives.
A trained staff of interviewers that is
geographically near the sample is
needed.
The total data collection period is
likely to be longer than for most
procedures.
Some samples (those in high-rise
buildings or high-crime areas, elites,
employees, students) may be more
accessible by some other mode.
Sampling limitations, especially as a
result of omitting those without
telephone.
Nonresponse associated with RDD
sampling is higher than with
interviews.
Questionnaire or measurement
constraints, including limits on
response alternatives, use of visual
aids, and interviewer observations.
Possibly less appropriate for personal
or sensitive questions if no prior
contact.
Especially careful questionnaire design
is needed.
Open questions usually are not useful.
Good reading and writing skills are
needed by respondents.
The interviewer is not present to
exercise quality control with respect to
answering all questions, meeting
questions objectives, or the quality of
answers provided.
Ineffective as a way of enlisting
cooperation (depending on group to
be studied).
Various disadvantages of not having
interviewer involved in data collection.
Need for good mailing addresses for
sample.
247
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
MARKETING RESEARCH IN ACTION 10-5
Hershey Research Sees Net Gain
Integrating old and new research without losing any is important to any company that puts work into
gathering the information. While moving toward ’Net research, firms are not going to throw away 10 years
of testing and start over from scratch. In 1999 and 2000, Hershey Food Corp. moved its all-important newproduct testing online, along with all of its historical testing data, in one king-sized commitment to ’Net
research. The move cut the time spent on the new product development process by two-thirds, a key strategic advantage in mature markets, and simultaneously ensured that a wealth of institutional knowledge
remained on hand even as the company’s roster of research professionals changed over the years.
While the company’s mail testing results might not come back for six weeks, online results can be garnered in two. Although Hershey took a giant step into online research, the company took small steps at
first to ensure that the transfer would go smoothly. The first task was validation testing. With more than
half of the U.S. population becoming more representative of the national population every day, Hershey
had to make sure that the results would be the same regardless of the method its researchers used. The
correlation between the two sets of tests was about 0.9 on a scale of 1.
With the online collection of data, Hershey has an easily accessible and searchable database. With this
new system there are fewer man-hours spent collecting the information, less human error, and greater
cost-effectiveness than with the old system. What Hershey has is a special application for companies
that wish to preserve their vast investment in past off-line research.
SOURCE: “Hershey Research Sees Net Gain,” Marketing News, November 25, 2002, p. 17.
and the administrator enters the sequence to start the interview. The interview then proceeds
in the same manner as the CATI described before, the only difference being that the information is input directly into the computer by the respondent. Computer interactive interviewing
has resulted in better responses from respondents and, in some instances, 30 to 40 percent cost
savings.26 Marketing Research in Action 10-5 gives an example of how and why Hershey converted its survey process to a computer interactive one—but without losing old research information. Thus Hershey succeeded in preserving its vast investment in past off-line research while
reaping the advantages of online methods.
Recent Developments in Software and Hardware.27 Many recent developments in both hardware and software have helped researchers to use computers more effectively. A few of them are
discussed here:
■ GRiDPAD is a pen computer that has an 8-in.-by-10-in. screen with a wand attached. Respondents use the wand to touch answer categories that appear on the screen (no more keyboards,
no more mice). Advanced Data Research (ADR) produces software that can be used with
GRiDPAD. Auto manufacturers were among the first to take advantage of the new system.
■ FormPro enables the researcher to create a custom-made, scannable questionnaire using standard paint and drawing tools. The finished questionnaire is then printed by a laser printer
using special scanning paper.
■ ScanPro reads forms from FormPro as well as those not made with FormPro. ScanPro, together
with an optical mark reader, enables the researcher to scan a questionnaire and produce a
standard ASCII file for later use with statistical software packages, spreadsheets, databases,
and so on.
■ ScanPhone is a high-tech telephone that enables consumers to scan and transmit bar codes.
This may offer new opportunities to researchers, because in the future even questionnaires
may be scanned.
248
Internet-Based Surveys. Although Web-based surveys have become quite popular and can be
advantageous, many problems remain, including choosing the right ’Net service provider and ’Net
software from a variety of choices available in the market. There is also the problem of the selection
bias, which occurs when the sample universe is comprised of individuals with the technical skills
COLLECTING DATA
to use the ’Net and the income to own a PC. This is not a random cross-section of the general
population. Another problem is attracting potential respondents to participate in the survey.
Some issues should be noted when conducting Internet-based surveys:
1. Check responses carefully. See if they are logically correct and acceptable; respondents may
provide false data.
2. Limit the number of characters allowed in the fill-ins.
3. Watch out for LANs. LAN servers may collect responses from several stations and send them
to Internet servers in bursts.
When these considerations have been taken care of, Internet-based surveys prove to be very
efficient and practical, and most important, they provide fast results.28 Marketing Research in
Action 10-6 gives an example of wireless marketing surveys conducted through cell phones
using Internet technology, which succeed in overcoming many limitations of traditional survey
modes.
Figure 10-2 and Table 10-4 illustrate previously described data collection methods with new
computerized methods.
Choice of Survey Methods for HMO Study
The HMO study described in earlier chapters successfully used a modification of the drop-off
method. The changes were made so that advantage could be taken of the availability of a complete address list of all faculty, staff, and married students. This population was scattered over
an area with a 20-mile radius, so it was not practical for the survey assistant to make unannounced personal visits. The specific procedure was as follows:
MARKETING RESEARCH IN ACTION 10-6
Cell Phones—Revolutionizing Mode
of Marketing Research
Study of consumer behaviors in their natural environment is the key to better consumer research.
Though very expensive, new research technologies and methods, such as the ethnographic method,
enable researchers to study consumers’ natural behavior, thus providing new insights to researchers.
Wireless marketing surveys address the expense issue while broadening and deepening the entire study
of consumer behavior.
Wireless marketing surveys are a data collection technique that uses standard wireless phones (cell
phones) to conduct marketing surveys either in text-based, voice-based, or a multimodal format. The
technical aspects of implementing these kind of surveys are delegated to specialized companies, while
the researchers focus on interaction with the respondents in real time.
The ability to deliver instant responses anywhere, anytime, and anyplace makes this technique
unique. At the same time, wireless surveys reflect consumer behavior better than other methods
because they record consumers’ experiences and feelings at the moment of purchase or consumption. This feature helps researchers overcome the weakness of retrospective methods, which rely on
the respondents’ biased and distorted memories. In addition, it not only helps respondents answer
such questions as “Why did I buy this product?” or “How do I feel about it?” (affective heuristic
factors) but also accurately measures customer satisfaction, interpersonal influences, and advertising effectiveness. This technique also helps researchers keep track of individual respondents (longitudinal behavior) and study location-sensitive topics in depth. From the consumers’ point of view,
wireless surveys are nonintrusive because they—to fill out the survey. All it takes is an Internetenabled cell phone, application service provider, and text-based short questions. Wireless surveys
will definitely open big doors in marketing research. Of course, the success of this mode will depend
on the cost of completed surveys versus the response rate.
SOURCE: “Calling Customers—Wireless Technology Ideal to Study Behavior,” Marketing News, January 20, 2003, p. 19.
249
Computer-Assisted Data Collection Methods
Personal
Fully
Automated
SelfInterviewing
✓
Allows respondents to choose own
schedule for completing survey
✓
✓
Web-Based
ComputerAssisted
Telephone
Interviewing
Fully
Automated
Telephone
Interviewing
✓
✓
Fully
Interactive
✓
Can easily incorporate complex
branching questions into survey
✓
✓
✓
✓
Can easily use respondent-generated
words in questions throughout the survey
✓
✓
✓
✓
Can accurately measure response times
of respondents to key questions
✓
✓
✓
✓
Can easily display a variety of graphics
and directly relate them to questions
✓
✓
✓
Eliminates need to encode data from
paper surveys
✓
✓
✓
✓
✓
✓
Errors in data less likely, compared to
equivalent manual method
✓
✓
✓
✓
✓
✓
Speedier data collection and encoding,
compared to equivalent manual method
✓
✓
✓
✓
✓
✓
FIGURE 10-2
Computer-assisted data collection methods.
SOURCE: Scott G. Dacko, “Data Should Not Be Manual Labor,” Marketing News, August 28, 1995, p. 31.
✓
✓ ✗
✓ ✗
✓
✓ ✗
✓ ✗
RESPONDENTS: SURVEY METHODS
Respondents need not have any
computer-related skills
ComputerAssisted
SelfInterviewing
Telephone
FROM
BENEFITS
ComputerAssisted
Personal
Interviewing
On-Site
CHAPTER 10 INFORMATION
250
SURVEYS
IN THE INTERNATIONAL
CONTEXT
TABLE 10-4
Eight Methods of Computerized Data Collection Available to Marketing
Researchers
Survey Methods
Characteristics
Computer-assisted personal interviewing
A method in which the researcher conducts in-person
interviews, reads questions to the respondent off a
computer screen, and keys the respondent’s answers
directly into the computer.
Computer-assisted self-interviewing
An on-site member of the research team intercepts and
directs willing respondents to nearby computers. Each
respondent reads questions off a computer screen and
keys his or her answers directly into a computer.
Fully automated self-interviewing
Respondents independently approach a centrally located
computer station or kiosk, read questions off a screen, and
key their answers directly into the station’s computer.
Computer-assisted telephone interviewing
Members of the research team telephone potential
respondents, ask questions of respondents from a
computer screen, and key the answers directly into a
computer.
Fully automated telephone interviewing
An automated voice asks questions over the telephone, and
respondents use keys on their touch-tone telephones to
enter their replies.
Web-based survey
Using batch type electronic mail, researchers send
emails to potential respondents with weblink. Potential
respondents go to the link that contains an interactive
survey. Respondents access the survey and enter their
answers.
1. A telephone contact was made with each household named in the sample, to determine the
head of the household and obtain that person’s agreement to participate. To minimize nonresponse bias, at least five telephone callbacks were made.
2. During the telephone call, arrangements were made to deliver the questionnaire, and a pickup time was set.
3. Before the pick-up, a reminder phone call was made to see if the questionnaire had been
completed.
4. If the questionnaire was not ready at the promised time, arrangements were made for a second pick-up time.
Only 3 percent of the sample of 1,500 persons refused to participate on the first contact. Fourteen percent were ineligible, not at home after calls, or had disconnected phones. Another 9 percent refused to complete the questionnaire they had received, usually citing lack of time (this is
not surprising, as the questionnaire had 14 pages and more than 200 variables). The overall
response rate of 74 percent reflected the interest of most respondents in the subject and the subtle
pressure to respond from the knowledge that someone was calling to pick up the questionnaire.
SURVEYS IN THE INTERNATIONAL CONTEXT
While conducting surveys for international research, a number of differences between the domestic and the international environment have to be taken into account. In this section the differences between domestic and international research and the problems faced by the researcher in
251
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
conducting international research are discussed briefly for the three major survey methods.
Often, surveys in the international context involve the use of translators and interpreters.
Correct use of the local language in a research instrument is critical to successful research design.
Owing to cross-national variances, the meaning of survey responses may get lost during translation and interpretation into other languages. Marketing Research in Action 10-7 provides a few
tips for effective interpretations in international surveys.
Personal Interviews
Personal interviewing tends to be the dominant mode of data collection outside the United
States.29 Lower wage costs imply that personal procedures are cheaper than in the United States.
On the other hand, use of personal interviewing requires the availability of field staff fluent in
the relevant language.
Often, however, given the lack of a pool of trained interviewers in other countries, companies with local research units or international research organizations may train and develop their
own field staffs. This provides greater control over the quality of the interviewing conducted in
different countries. This is in marked contrast to the practice of “buying field and tab services”
from an outside organization that is common in the United States. These interviewers are not
necessarily required to work exclusively for a given research supplier, though often they may do
so of their own choice.
MARKETING RESEARCH IN ACTION 10-7
Tips for Effective Interpretations
Language is perhaps the most important barrier when researching international markets. Therefore,
irrespective of the research instrument being used, ensuring exact interpretations is vital in deriving
insightful results. The following tips would be of help to researchers using interpreters.
1. An interpreter is different from a translator!
While it may seem obvious, people generally use one for the other. Therefore, it is critical to understand their terminologies. A translator is someone who works with the written word and translates
a written text in one language into a written text in another language. In contrast, an interpreter
works with the spoken word or oral communication, and is used for real-time communication in a
foreign country.
2. It’s all about the meaning
Interpreters do not translate. Rather, they communicate what researchers mean in the listener’s
language. There are plenty of examples of marketing gaffes that have relied heavily on literal
translations and have produced erroneous and often embarrassing results. Therefore, it is not about
the literal translation—it’s the appropriate rendering of your meaning in the other language.
3. A pause goes a long way
If you speak for too long without a pause, the interpreter will find it more difficult to render your
meaning and your listeners will get bored or lose focus. Therefore, a pause helps the interpreter to
collect, assimilate and effectively interpret the message into the foreign language.
4. Go easy on the slangs
The focus should be on the message that has to be communicated. This calls for easy and straight
talk that does not involve slangs, humor, pun, or play with words. This helps the interpreter to concentrate on interpreting the message correctly.
5. Look at your counterpart, not the interpreter!
In the most ideal setting, the interpreter should be invisible, with only the interpreted messages
going back and forth. This is possible only if the researcher focuses his attention on the counterpart,
and using the ears only with the interpreter. Such an approach also helps the researcher to pick up
the nonverbal cues from the counterpart.
SOURCE: Anne Orban, “Qualitatively Speaking: Ten tips for using interpreters in international research,” Quirk’s Marketing Research Review, November 2005,
p. 20. http://www.quirks.com/articles/a2005/20051112.aspx?searchID-249352105.
252
SURVEYS
IN THE INTERNATIONAL
CONTEXT
The ease with which the cooperation of respondents can be obtained may vary, however,
from one country or culture to another. In Latin countries, and particularly in the Middle East,
interviewers are regarded with considerable suspicion. In Latin countries, where tax evasion is
more prevalent, interviewers are often suspected of being tax inspectors. In the Middle East,
where interviewers are invariably male, interviews with housewives often have to be conducted
in the evenings when husbands are at home.
Telephone Interviews
In international marketing research, the advantages of these telephone interviews are not always
as evident. Low levels of telephone ownership and poor communications in many countries limit
the coverage provided by telephone surveys. In addition, telephone costs are often high, and volume rates may not be available. Again, this depends on the specific country and the target population. Consequently, the desirability of conducting a telephone survey will depend to a large
extent on the nature and purpose of the survey.
In industrial international marketing research, the use of telephone surveys may be quite
effective. Most businesses, other than some small or itinerant retailers or craftspersons, are likely
to have telephones.
With the decline of international telephone costs, multicountry studies can also be conducted
from a single location. This significantly reduces the time and costs associated with negotiating
and organizing a research project in each country, establishing quality controls, and so on.
Although the additional costs of making international telephone calls are incurred, these may not
be highly significant when a centralized location is used.30
International calls also obtain a higher response rate. Results obtained using this technique
have been found to be highly stable. Interviewer and client control is considerably greater. The
questionnaire can be changed and modified in the course of the survey, and interviewing can be
extended or stopped to meet the client’s requirements. It is necessary to find interviewers fluent
in the relevant languages, but in most European countries this is rarely a problem.
In consumer research, the feasibility of using telephone surveys depends on the level of private telephone ownership in a country, and the specific target population. In countries such as
India, telephone penetration is very low. According to the Ministry of Communications and Information Technology in 2004, India’s rural telephone penetration was 4.3 main lines per 100 populations and urban telephone density was 15.2 per 100 populations.31 Even in relatively affluent
societies such as Great Britain, telephone penetration is only 80 percent, and telephone interviewing is not widely used because many practitioners are still skeptical about it. In Britain and
France, there are substantial declines in telephone response rates in large cities. However, these
declines are not observed in Germany and Switzerland. In some of the Eastern European countries and some members of the Eastern European Commonwealth of Independent States which
have poor telecommunication systems, conducting telephone surveys may not be a good idea.
Self-Administered Surveys
As in the case of telephone interviews, the advantages and limitations of self-administered surveys in international marketing research are not always clear, because of the absence of mailing
lists, poor mail services, limited email use, limited web access, and high levels of illiteracy. In
many markets the efficacy of self-administered surveys depends on the specific product market
being investigated (that is, industrial versus consumer), and also the nature of the survey. For
example, response rates are a problem in international mail surveys. When international mail
surveys are conducted, the response rates vary depending on the country to which the questionnaires are mailed. The response rate for a mail survey conducted by ESOMAR in 1992 is
given in Table 10-5, but these rates should not be viewed as representative of the country’s
response rates.
253
CHAPTER 10 INFORMATION
FROM
RESPONDENTS: SURVEY METHODS
TABLE 10-5
Response Rates in the ESOMAR 1992 Price Study
Sample
Selected
Effective
Returns
Percent of Response
North America
22
11
50.00
Latin America
17
16
94.12
Eastern Europe
17
10
58.82
Turkey
11
10
90.91
North Africa
6
4
66.67
Middle East
9
6
66.67
Australasia
9
6
66.67
India
5
4
80.00
Pacific Rim
12
11
91.67
Japan
12
5
41.67
South Africa
Total
3
3
100.00
123
86
Overall ⫽ 69.92
SOURCE: European Society for Opinion and Marketing Research Prices Study, 1992.
Self-administered surveys typically can be used effectively in industrial international marketing research. Mail and email lists such as those from Bottin International, or directories for
specific industries, are generally available.
In consumer research, and particularly in developing countries, the use of self-administered surveys may give rise to some problems. Mailing addresses and email lists comparable to those in the
domestic market may not be available, or not sold, and public sources such as telephone directories may not provide adequate coverage. Lists that are available, that is, magazine subscription lists
or membership association lists, may be skewed to better-educated segments of the population. In
addition, in some countries, the effectiveness of self-administered surveys is limited not only by low
literacy levels but also by the reluctance of respondents to respond to them. As noted previously, levels
of literacy are often less than 40 or 50 percent in some Asian and African markets, thus limiting the
population that can be reached by mail or email. Even in countries where literacy levels and mail services make the use of mail surveys feasible, a tendency to regard surveys as an invasion of privacy
may limit their effectiveness.
Thus, while mail surveys may be used effectively in industrial marketing research, in consumer
research they may be appropriate only in industrialized countries where levels of literacy are high
and mailing lists are generally available.
Summary
The choice of a data collection method involves a series of compromises in matching the often conflicting
requirements of the situation with the strengths and limitations of the available methods. While each situation is unique to some degree, the following represent the major constraints to be satisfied:
Available budget
Nature of the problem
Need for accuracy
The complexity of the required information
How quickly the results are needed
254
QUESTIONS
AND
PROBLEMS
Part of the skill of research design is adapting a basic data collection method, whether personal or telephone interview or self-administered survey, to those constraints. This process of adaptation means exploiting the advantages as well as blunting the limitations, which to some extent makes generalizations suspect.
Nonetheless, it is useful to summarize the relative merits of the basic survey methods, as in Table 10-3, to put
the methods into perspective. Of course, these summary judgments do not reflect the myriad special factors
that may be influential in specific cases, such as the availability of a sampling frame, the need to ask sensitive questions, and the rarity of the population to be sampled.
The greatest potential for effective adaptation of method to situation lies with combinations of methods,
or specialized variants of basic methods. The latter include omnibus surveys and mail panels, which are particularly good for tracking studies, or where limited responses from specific populations are required and
budget constraints are severe. Some combinations of methods, such as the drop-off method, similarly confer
impressive advantages by combining some of the advantages of personal interviews and mail surveys. As
with specialized variants, however, their usefulness often is restricted to certain settings.
Survey research is bound to be the most dominant data collection method in the future. The future of
survey research is exciting. With technology growing at such a tremendous pace, newer and better methods
to conduct surveys, such as web-based surveying and email surveys, are becoming more common. Another
trend in the survey research area is internationalization. This introduces a number of new challenges to the
marketing researcher, who has to contend with language, cultural, and a host of other problems.
Questions and Problems
1. What kind of data collection procedure would you
recommend to research the question of why female
shoppers choose a particular retail store at which
to buy clothing?
2. Even with the use of expensive gifts and callbacks,
all three interview methods produce low response
rates, usually under 30 percent. What are some
ways to increase these response rates?
3. What biases, if any, might be introduced by offering to give respondents $5.00 when they return a
mail questionnaire? Would these biases be different if a gift with a retail value of $5.00 were
included with the questionnaire?
4. What can an interviewer do during the first 30 seconds of an interview to maximize the cooperation
rate of (a) a personal interview and (b) a telephone
interview? Assume the appropriate respondent is
the head of the household.
5. What are the advantages of the telephone prenotification approach over a conventional mail
survey?
6. Write a cover letter for a mail survey of householders that contains a number of questions regarding
the utilization of various kinds of burglaryprotection devices.
7. You are a senior analyst in the marketing research
department of a major chemical company. Your
company has accidentally stumbled upon a chemical that, when combined with plastic, gives it near
metallic properties. You have been asked to find
8.
9.
10.
11.
out various uses for this chemical and also forecast
its total market potential.
a. What information if any, that could be obtained
from respondents would be useful for this
research?
b. What techniques are applicable for obtaining
each item of information?
c. Design a survey to obtain the information desired. Prepare all instructions, collection forms,
and other materials required to obtain such
information.
d. Estimate the cost of conducting the survey you
have designed.
You are a publisher of national repute and you
would like to find the potential for an academic
book. You have decided to conduct a survey
among college professors for this purpose. What
form of survey method would you use? What
trend do you see in the future for this survey
method?
What are the advantages and disadvantages of
computer interactive interviewing compared to the
traditional methods of surveying?
What are the problems faced by researchers when
they are conducting research in developing countries?
You are the manufacturer of a major consumer
product in the United States. You plan to enter the
following countries with your products within the
next two years: Russia, France, South Africa,
Brazil, China, Japan, Mexico, Canada, India, and
255
Germany. Based on the demographic data and the
infrastructural data about these countries, determine the best survey method to be adopted in each
of these countries.
12. Idaho Ideal Potatoes, Ltd., developed a new
guacamole-flavored potato chip and wanted to
ascertain the national sales potential for this unique
product offering. However, the market researcher
found the mail response rate to be extremely low
(10 percent) and that many of the returned surveys
had been completed incorrectly.
a. How might the company have collected the
relevant data more effectively?
b. What are the advantages and disadvantages of
using telephone surveys and focus groups for
this kind of study? Due to stringent budget
constraints, the marketing research director
decided to limit the survey to residents of San
Antonio, Texas.
c. What possible biases could be reflected in the
survey results?
d. Idaho Ideal Potatoes, Ltd., has decided to offer
their new potato chip simultaneously in foreign
markets and wishes to survey potential sales
demand for the product in the United Kingdom.
How should the choice of survey method be
altered to suit the company’s research of foreign
markets?
13. Judy Gomez, head of marketing for Wildlife
Treasures, Inc., has been informed by the board of
directors that they would like a marketing
research study conducted to determine the cause
of the plateau in sales performance over the past
two years. The company is a producer of art and
ornaments depicting African wildlife, and currently supplies specialty stores throughout the
United States. A considerably restricted budget
has been allocated to the project. Ms. Gomez is
expected to survey all of these stores in order to
determine how sales in the domestic market can
be increased. The board of directors has been
generous in its time allowance for the study,
preferring a thorough marketing research effort
to a hasty one.
a. Suggest a suitable survey method for the study
and justify your answer. Ms. Gomez believes that
a possible remedy to the company’s stagnant
sales performance may be an entry into foreign
markets. Because Wildlife Treasures has received
some unsolicited inquiries from European
tourists, she feels that a market may exist in
Europe for its specialty goods.
b. Given the restraints of the study, how can
Ms. Gomez conduct a survey to determine
potential markets across Europe?
c. Ms. Gomez would like to utilize the company’s
recently updated computer system for the
marketing research study. What features would
need to be available in order to use the computer system for this purpose?
14. A museum of arts in a large city wants to conduct
a study of museum patrons’ attitudes concerning
its collection of exhibits. What survey method
would you suggest? Why?
End Notes
1. Steve Lewis, “Fear Not, We Come in Peace: You
Can Convert Survey-Phobics into Lasting Relationships,” Marketing News, January 4, 1999, Vol. 33,
No. 1, p. 18.
2. Seymour Sudman, “Improving the Quality of
Shopping Center Sampling,” Journal of Marketing
Research, 17, November 1980, pp. 423–431.
3. Frederick Wiseman, Marianne Schafer, and Richard
Schafer, “An Experimental Test of the Effects of a
Monetary Incentive on Cooperation Rates and
Data Collection Costs in Central-Location Interviewing,” Journal of Marketing Research, November
1983, pp. 439–442.
4. S. M. McIntyre and S. D. F. G. Bender, “The Purchase Intercept Technique in Comparison to Telephone and Mail Surveys,” Journal of Retailing, 62,
Winter 1986, pp. 364–383.
5. Ellen Gregory, “Cost/Quality Issues Plague Mall
Intercepts,” Marketing Research, Summer 1996,
pp. 46–47.
256
6. Nanci A. Glassman and Myron Glassman, “Screening Questions: How Screening Questions Can
Cause Self-Selection Bias,” Marketing Research, Fall
1998, pp. 26–31.
7. “The Frame,” published by Survey Sampling, Inc.,
September 1996.
8. Stanley L. Payne, “Data Collection Methods: Telephone Surveys,” in Robert Ferber, ed., Handbook of
Marketing Research. New York: McGraw-Hill, 1974,
pp. 2.105–2.123.
9. Payne, Handbook of Marketing Research. New York:
McGraw-Hill, 1974, p. 2.116.
10. J. B. Lansing and J. N. Morgan, Economic Survey
Methods, Ann Arbor, MI: University of Michigan,
Institute of Social Research, 1971, pp. 203–243.
11. “So how good is RDD?” Marketing Research, Fall
2007, pg. 5.
12. Quoted from Census Bureau sources by Paul
L. Erdos, “Data Collection Methods: Mail
Surveys,” in Robert Ferber, ed., Handbook of
13.
14.
15.
16.
17.
18.
19.
20.
Marketing Research, New York: McGraw-Hill, 1974,
pp. 2–91.
Results from other comparative studies have not
been so clear-cut. See, for example, W. Locander,
S. Sudman, and N. Bradburn, “An Investigation
of Interviewer Method, Threat and Response
Distortion,” Journal of the American Statistical Association, June 1976, pp. 262–275.
Michael J. Houston and Neil M. Ford, “Broadening
the Scope of Methodological Research on Mail
Surveys,” Journal of Marketing Research, 13, November 1976, pp. 397–403.
Quoted from Census Bureau by Paul L. Erdos,
“Data Collection Methods: Mail Surveys,” in
Robert Ferber, ed., Handbook of Marketing Research,
New York: McGraw-Hill, 1974, pp. 2–91.
Council of American Survey Research Organizations, “On the Definition of Response Rates,” Special Report, Port Jefferson, NY: CASRO, 1982.
Frederick Wiseman and Maryann Billington,
“Comment on a Standard Definition of Response
Rates,” Journal of Marketing Research, August 1984,
pp. 336–338.
Leslie Kanuk and Conrad Berenson, “Mail Surveys and Response Rates: A Literature Review,”
Journal of Marketing Research, 12, November 1975,
pp. 440–453; Jeanine M. James and Richard
Bolstein, “The Effect of Monetary Incentives and
Follow-up Mailings on the Response Rate and
Response Quality in Mail Surveys,” Public Opinion
Quarterly, 54, Fall 1990, pp. 346–361.
J. Scott Armstrong and Terry S. Overton,
“Estimating Nonresponse Bias in Mail Surveys,”
Journal of Marketing Research, 14, August 1977,
pp. 396–402.
Dick McCullough, “Web-based Market Research
Ushers in New Age,” Marketing News, September
14, 1998.
CASE 10-1
Roland Development Corp.
Roland Development was a leading builder of homes
in the western United States. Its emphasis was on condominiums and ownhouses, which were forecast to
have an attractive future in these markets. These housing types lent themselves to standardization and costreduction possibilities. Further, rising land costs were
causing the share of single-family detached houses
to decline significantly. Meanwhile, the share of market
for single-family attached houses (houses with common
21. Nina M. Ray and Sharon W. Tabor, “Cyber Surveys,”
Marketing Research, Spring 2003, pp. 35–37.
22. Marvin A. Jolson, “How to Double or Triple MailSurvey Response Rates,” Journal of Marketing, 41,
October 1977, pp. 78–81.
23. David Schwartz, “Locked Box Contains Survey
Methods, Helps End Some Woes of Probing
Industrial Field,” Marketing News, January 27,
1978, p. 18.
24. Christopher H. Lovelock, Ronald Stiff, David
Cullwick, and Ira M. Kaufman, “An Evaluation of
the Effectiveness of Drop-off Questionnaire Delivery,” Journal of Marketing Research, 13, November
1976, pp. 358–364.
25. Gail Gaboda, “Phone Cards Help Researchers
Reward Respondents,” Marketing News, September
15, 1997, p. 14.
26. For more details on cost savings, refer to John P.
Liefeld, “Response Effects in Computer Administered Questioning,” Journal of Marketing Research,
25, November 1988, pp. 405–409.
27. Lewis C. Winters, “Questionnaires in the 1990s:
Wands and Scannable Forms Are ‘In’,” Marketing
Research, June 1992, p. 46.
28. Robert Hays, “Phone Cards Help Researchers
Reward Respondents,” Marketing News, April 13,
1998, p. 13.
29. D. Monk, “Marketing Research in Canada,”
European Research, November 1987, p. 274; J. J.
Honomichl, “Survey Results Positive,” Advertising
Age, November 1984, p. 23.
30. M. De Houd, “International Computerized Telephone Research: Is It Fiction?,” Marketing Research
Society Newsletter, January 1982, pp. 14–15.
31. ”Doing ebusiness in India,” The Economist Intelligence Unit, The Economist, http://globaltechforum.
eiu.com/index.asp?layout-newdebi&country_id=
IN&title=Doing+ebusniess+in+India.
walls, floors, or roofs) was expected to double in the
next five years. Roland was well positioned to exploit
these trends by following a strategy that differed from
competition in three areas:
■ Market segmentation. Roland typically segmented
the market more finely than other home builders,
and then designed homes to meet the specific needs
of these groups.
■ Direct selling. Shoppers in some department stores
could find full-scale, fully furnished Roland homes
on display.
257
■ Low prices for a complete housing package
(including all the furnishings and necessary
financing).
The company had begun to expand its limited
line of condominiums and townhouses to provide
design and square-footage combinations that would
appeal to higher-income households. The management was especially pleased with the elegance, convenience, and durability of the four new models they
were planning to launch. Several problems remained
to be solved. The first was the identification of a creative strategy that would position the new models and
attract the largest number of purchases. That is, the
company wanted to know what main ideas and
themes should be used in the advertising of the new
models. Another problem was to identify those segments of the market with the highest probability of
purchasing the new models. The company asked the
YKG Group, a large national research firm, to submit
a written proposal for research which would provide
Roland’s management with information useful in
solving these two problems. Their proposal is summarized below.
Research Proposal
The recommended research design would use a consumer panel and employ both telephone interviews
and mail questionnaires. The research firm felt that the
needed information could be obtained only from that
very small proportion of the population who might
buy such a home. Each of several different market segments would be studied to determine how they positioned the new models in relation to competing homes
already on the market. The likelihood of purchasing a
Roland model would also be determined during the
study for each of the three market segments and also
for each of several different advertising themes. This
information would help Roland identify the most
promising market segments for the new models, as
well as the creative advertising strategy that would
most appeal to them.
The proposed research design consisted of three
phases: (1) The members of a large consumer mail
panel would be screened to locate qualified prospects
for the new models; (2) a relatively small sample of
qualified prospects would be interviewed “in depth”
to identify possible advertising themes; and (3) a
large sample of qualified respondents would be surveyed by mail to test their response to alternative creative strategies.
Phase 1. The YKG Group maintained a bank of over
200,000 families who agreed to cooperate in research
projects undertaken by the firm. Considerable information existed about each family, including geographic
location, occupation and age of male and female heads
of family, total family income, presence and age of
258
children, and so on. Roland managers felt that the four
new models would most likely appeal to middle- and
upper-income families of size two, three, or four, with
a house-hold head 30 years of age or over. For this reason, the first phase of the proposed research involved
mailing a short questionnaire to all panel members
with those characteristics. The questionnaire asked
panel members to indicate the likelihood of their purchasing a home in the next two to three years and also
to report their attitude toward buying a townhouse or
condominium.
It was expected that this screening process
would locate some 3,000 to 5,000 families who would
be prospects to buy the new models over the next
few years. To be considered a prospect, a family had
to report being likely to purchase a home in the next
two to three years, as well as having a favorable attitude toward a condominium or townhouse. Among
these prospects, three market segments would be
identified. A high-income family would be a “very
good” prospect if it was “very likely” to buy a home;
a medium-income family would be a “good”
prospect if it was “very likely” to buy a new home;
and a high-income family would be a “fair” prospect
if it was “somewhat” likely to buy a new home. All
other responses were considered to indicate nonprospects.
Phase 2. In this phase about 200 qualified
prospects would be interviewed using a combination
of telephone and mail. These families would be
mailed pictures, specifications, and line drawings of
the company’s new models of condos and townhouses, although they would not be identified by the
Roland name. The line drawings would include front
and rear views of each unit’s exterior as well as
sketches of each room. The specifications would
include the number of square feet, wall thickness,
heating and cooling equipment capacities, appliance
brands and models, slab thickness, type of roof covering, and other features.
After reviewing these materials, respondents’
reactions and impressions would be obtained through
telephone interviews using open-ended questions.
Interviewers would be told that the objective was to
obtain qualitative data useful for ascertaining how
potential buyers perceived the new models with
respect to appearance, comfort, elegance, convenience,
durability, ease and economy of maintenance, and
other criteria. Interviewers would be instructed to
record verbatim responses and were told that it was
very important to do so because none of the responses
would be tabulated or analyzed statistically. Responses
to the open questions would then be studied to identify four or five ideas of themes that might be considered for use as creative strategies in advertising the
new models.
Phase 3. This phase would be undertaken after four
of the best advertising themes had been identified.
Some 2,400 families would be selected from the list of
prospects obtained from Phase 1—approximately 800
“very good” prospects, 800 “good” prospects, and 800
“fair” prospects.
All of the families in each of the three market segments would be sent pictures, line drawings, and
specifications (including prices) of the new Roland
models as well as those of major competing models,
all identified by brand name. Each of these three
groups of prospects would then be randomly divided
into four subsamples of 200, each of which would
receive one, and only one, of the four advertising
themes identified for the new models. Thus, the study
design would consist of three samples of 800 families
each. In turn, each sample would be broken into four
subsamples, each of which would receive a different
advertising theme.
Analysis. The effect of each advertising theme on
each prospect segment would be evaluated on three
measures: the degree to which it (1) resulted in the
new line being rated as “most appealing,” (2) led
respondents to request further information about the
company’s products, and (3) led respondents to indicate that they would be most likely to select one of
the company’s homes if they were to make such a
purchase in the near future. For each advertising
theme-prospect segment combination the research
would yield three percentages. For example, for
theme #1 and the “very good” prospects, the research
might show that 38 percent of the respondents found
a model in the new line “most appealing” among all
the models reviewed; that 26 percent requested further information about the Roland models; and that
17 percent indicated that they “most likely would
purchase” one of the new Roland models. By comparing these three percentages for each advertising
theme-prospect segment combination, it would be
possible to identify the most promising combinations.
These results could be weighed by the relative size of
each prospect segment to decide which creative strategy would be most effective in generating sales interest in the new models.
Questions for Discussion
1.
2.
Would you recommend that Roland accept the
YKG Group proposal?
If yes, what conclusions can be drawn from the
data in Phase 3 of the research?
3.
If the proposal is not accepted, what alternative
designs should be considered?
259
11
ATTITUDE MEASUREMENT
Learning Objectives
■ Introduce the concept of measurement and scaling in marketing research.
■ Briefly discuss the different scales in measurement.
■ Introduce the different types of scales used for measuring attitudes.
■ Give a description of each of the well-known scales that are used to measure attitudes.
■ Provide an approximate heuristic for choosing an attitude scale.
■ Discuss the concepts of reliability, validity, and generalizability.
Most questions in marketing research surveys are designed to measure attitudes. The answers to
these questions help decision makers in designing specific products or services to cater to the
needs of segmented audiences. For example, each of the following situations involves the measurement of some aspect of a respondent’s attitude:
1. An appliance manufacturer wants to know how many potential buyers are aware of a brand
name. (What brand names do they think of in connection with dishwashers?)
2. Administrators concerned with formulating an energy policy want to know what proportion
of voters agree that car buyers should pay an extra tax of several hundred dollars on cars that
get poor gasoline mileage.
3. A food manufacturer is interested in the intentions of a sample of consumers to buy a possible new product after the concept has been described to them.
Common to each of these examples is a need to learn something about the basic orientation or
attitude of present or prospective customers. Their attitudes are based on the information they
have, their feelings (liking and disliking), and their intended behavior.
What management really wants to understand—and ultimately influence—is behavior. For
many reasons, however, they are likely to use attitude measures instead of behavior. First, there
is a widely held belief that attitudes are precursors of behavior. If consumers like a brand, they
are more likely to choose that brand over one they like less. Second, it is generally more feasible
to ask attitude questions than to observe and interpret actual behavior. Attitude measures offer
the greatest advantage over behavior measures in their capacity for diagnosis or explanation. Attitude measures can be used to help learn which features of a new product concept are acceptable
or unacceptable, as well as the perceived strengths and weaknesses of competitive alternatives.
Insights can be gained into the process by which choice decisions are made: What alternatives are
known and considered? Why are some rejected? What problems are encountered with the products or services that are used?
This chapter is concerned primarily with the measurement of attitudes. Some measurement
approaches were encountered in earlier chapters. Projective techniques and physiological methods, discussed in Chapter 8, are indirect methods for inferring a person’s attitude. By far the
most popular approach is the direct self-report, in which the respondent is asked a series of questions. The two previous chapters described the survey methods appropriate for such self-reports.
260
WHAT
ARE
ATTITUDES?
This chapter and the next are devoted specifically to attitude measurements, in recognition of
their importance to marketing and of the special problems of specifying and identifying attitudes.
WHAT ARE ATTITUDES?
What is attitude? How can it be measured for marketing research purposes? Attitude is psychological constructs, a way of conceptualizing the intangible. Attitudes can’t really be observed or
measured directly because their existence is inferred from their consequences. On the other hand,
people’s values and beliefs may dictate or affect their purchasing decisions. Values and beliefs
in retrospect are influenced by a person’s attitude; conversely, values are the determinants of attitudes and belief involves evaluation.
A person with more positive than negative beliefs toward a psychological object is judged
to have a positive attitude. Attitude measurement techniques are generally systematic methods for abstracting the effective component of belief systems in order to generate an attitude
score. Attitudes are the essence of the “human change agent” that all marketers strive to influence, but without the right tools to effectively measure attitude, attitudinal research has little
to offer.1
Attitudes are mental states used by individuals to structure the way they perceive their environment and guide the way they respond to it. There is general acceptance that there are three
related components that form an attitude: a cognitive or knowledge component, a liking or affective component, and an intentions or actions component. Each component provides a different
insight into a person’s attitude.
Cognitive or Knowledge Component
The cognitive or knowledge component represents a person’s information about an object. This
information includes awareness of the existence of the object, beliefs about the characteristics or
attributes of the object, and judgments about the relative importance of each of the attributes.
Consider the knowledge people might bring to planning a ski vacation in the Rockies. They
might remember the names of several ski areas without prompting: Aspen, Snowmass, Alta, and
Park City, for example. This is unaided-recall awareness. The names of additional ski areas are
likely to be remembered when the travel agent mentions them. This is aided-recall awareness.
Knowledge of ski areas is not limited to awareness, however. From the experience of
friends, brochures, magazine articles, and other sources, a person will have formed beliefs or
judgments about the characteristics or attributes of each of these ski areas. These attributes
might range from the difficulty of the slopes to the type of social life and the cost of accommodations. Often, these beliefs incorporate explicit comparative judgments within a set, such
as which ski area is the most difficult or the cheapest. Another kind of belief is an overall similarity judgment: Are Aspen and Snowmass more similar to each other than Aspen and Alta,
for example?
Affective or Liking Component
The affective or liking component summarizes a person’s overall feelings toward an object,
situation, or person, on a scale of like–dislike or favorable–unfavorable. When there are several
alternatives to choose among, liking is expressed in terms of preference for one alternative
over another. Preferences can be measured by asking which is “most preferred” or the “first
choice,” which is the “second choice,” and so forth. Affective judgments also can be made
about the attributes of an object, such as a ski area. Someone may like all other aspects of an
area but dislike the location because it requires too much traveling.
261
CHAPTER 11 ATTITUDE MEASUREMENT
Intention or Action Component
The intention or action component refers to a person’s expectations of future behavior toward
an object. Is he or she “very,” “somewhat,” or “not at all” likely to go to Aspen for a ski week
next winter? Intentions usually are limited to a distinct time period that depends on buying
habits and planning horizons. The great advantage of an intentions question is that it incorporates information about a respondent’s ability or willingness to pay for the object, or otherwise
take action. One may prefer Aspen over all other ski areas in the Rockies but have no intention
of going next year because of the cost.
THE CONCEPT OF MEASUREMENT
AND SCALING
Measurement can be defined as a standardized process of assigning numbers or other symbols to certain characteristics of the objects of interest, according to some prespecified rules.
Measurement often deals with numbers, because mathematical and statistical analyses can be
performed only on numbers, and they can be communicated throughout the world in the
same form without any translation problems. For a measurement process to be a standardized process of assignment, two characteristics are necessary. First, there must be one-to-one
correspondence between the symbol and the characteristic in the object that is being measured. Second, the rules for assignment must be invariant over time and the objects being
measured.
Scaling is the process of creating a continuum on which objects are located according to the
amount of the measured characteristic they possess. An illustration of a scale that is often used
in research is the dichotomous scale for sex. The object with male (or female) characteristics is
assigned the number 1 and the object with the opposite characteristics is assigned the number 0.
This scale meets the requirements of the measurement process in that the assignment is one to
one and it is invariant with respect to time and object. Measurement and scaling are basic tools
used in the scientific method and are used in almost every marketing research situation. The
decision of what type of scale is critically important since the scale type will dictate the
method(s) of analysis that you will be able to use to analyze the data collected. For example, if
you want to be able to calculate the average satisfaction the scale for the question asked must
have certain scale properties be either inverval or ratio. If you want to show that two constructs
are correlated they each must be measured with an interval or ratio scale. Therefore, you should
think about the analysis you want to perform as you design the scales you will use on your
questionnaire.
Properties of Measurement Scales
262
The assignment of numbers is made according to rules that should correspond to the properties
of whatever is being measured. The rule may be very simple, as when a bus route is given a number to distinguish it from other routes. Here, the only property is identity, and any comparisons
of numbers are meaningless. This is a nominal scale. At the other extreme is the ratio scale, which
has very rigorous properties. In between the extremes are ordinal scales and interval scales, as
shown in Table 11-1.
Attitude variables, such as beliefs, preferences, and intentions, are also measured using rating
scales. These scales provide respondents with a set of numbered categories that represent the range
of possible judgments or positions. An attitude scale involves measurement in the same sense that
a thermometer measures temperature or a ruler measures distance. In each of these cases,
measurement means the assignment of numbers to objects or persons to represent quantities of
their attributes. For example, the attributes of a person include his or her income, social class, attitude, and so forth. Therefore, it is very important to understand the differences among the types
THE CONCEPT
OF
MEASUREMENT
AND
SCALING
TABLE 11-1
Types of Scales and Their Properties
Types of
Measurement
Scale
Types of
Attitude Scale
Rules for
Assigning Number
Other Typical
Marketing
Applications
Nominal
Dichotomous “yes” or
“no” scale; “buy” or
“not buy” scale
Objects are
either identical or
different
Classification (by
sex, geographic
area, social class)
Percentages,
mode/chi-square
Ordinal or
rank order
Comparative, rank
order, itemized
category, paired
comparison
Objects are
greater or smaller
Rankings
(preference, class
standing)
Percentile, median,
rank-order
correlation/
Friedman ANOVA
Interval
Likert, Thurstone,
Stapel, associative,
semantic-differential
Intervals between
adjacent objects
are equal
Index numbers,
temperature
scales, attitude
measures
Mean, standard
deviation, product
moment correlations/
t-tests, ANOVA, regression, factor analysis
Ratio
Certain scales with
special instructions
such as a constant
sum scale
There is a meaningful zero, so
comparison of
absolute magnitudes between
objects is possible
Sales, incomes,
units produced,
costs, age
Geometric and
harmonic mean,
coefficient of variation
Statistics/
Statistical Tests
of scales and to be able to identify them in practice, for their properties put significant restrictions
on the interpretation and use of the resulting measurements.
Nominal Scale. In a nominal scale, objects are assigned to mutually exclusive, labeled categories,
but there are no necessary relationships among the categories; that is, no ordering or spacing is
implied. If one entity is assigned the same number as another, they are identical with respect to a
nominal variable. Otherwise, they are just different. Sex, zip code, occupation, and marital status
are nominally scaled variables. The only arithmetic operation that can be performed on such a scale
is a count of each category. Thus, we can count the number of automobile dealers in the state of
California or the number of males seen on a given bus route in the past hour.
Ordinal Scale. An ordinal scale is obtained when objects are ranked or arranged in order with
regard to some common variable. The question is simply whether each object has more or less of
this variable than some other object. The scale provides no information as to how much difference
there is between the objects.
Because we do not know the amount of difference between objects, the permissible arithmetic operations are limited to statistics such as the median or mode (but not the mean). For
example, suppose a sample of 1,000 consumers ranked five brands of frozen mixed vegetables
according to quality. The results for Birds Eye brand were as follows:
Quality Rank
Number of Respondents Giving
Ranking to Birds Eye
First
150
Second
300
Third
250
Fourth
200
Fifth
100
Total
1,000
263
CHAPTER 11 ATTITUDE MEASUREMENT
The “second” quality category is the mode; the “third” category is the median; however, it is
not possible to compute a mean ranking, because the differences between ordinal scaled values
are not necessarily the same. The finishing order in a horse race illustrates this type of scale since
while one would know the order they finished, the distance between the horses could be very
different. Similarly, brands of frozen vegetables can be ranked according to quality, from highest
to lowest.
Interval Scale. In an interval scale the numbers used to rank the objects also represent equal increments of the attribute being measured. This means that differences can be compared. The difference between 1 and 2 is the same as between 2 and 3, but is only half the difference between 2
and 4. The location of the zero point is not fixed, and therefore zero does not denote the absence
of the attribute. Fahrenheit and Celsius temperatures are measured with different interval scales
and have different zero points. Interval scales have very desirable properties, because virtually the
entire range of statistical operations can be employed to analyze the resulting number, including
addition and subtraction. Consequently, it is possible to compute an arithmetic mean from intervalscale measures.
A recurring question regarding most attitude measures is whether or not they are interval
scales. Usually it is doubtful that the intervals between categories are exactly equal, but they may
not be so different as to preclude treating the whole as an interval scale. A good example is a
“willingness to buy” scale with 10 categories labeled from 1 to 10. If this were an interval scale,
we could say that two people with scores of 2 and 4, respectively, differed by the same degree
of “willingness” as two other people with scores of 8 and 10. Further, only ratios of the differences in scale values can be meaningfully interpreted, not ratios of the absolute scale values. For
example, the difference between 8 and 10 is twice the difference between 2 and 3, but 6 on the
“willingness” scale does not represent three times the value of 2 in terms of the degree of willingness. This is due to the fact that zero has no real meaning.
Ratio Scale. A ratio scale is a special kind of interval scale that has a naturally defined zero
point. Examples would include weight, market share, or dollars in savings accounts. With a
ratio scale it is possible to say how many times greater or smaller one object is than another.
This is the only type of scale that permits us to make comparisons of absolute magnitude. For
example, we can say that an annual income of $80,000 is two times as large as an income of
$40,000.
There have been some contemporary efforts to adapt ratio scales to the measurement of social
opinion. Some researchers have attempted to use magnitude estimation scales to overcome the
loss of information that results when categories arbitrarily constrain the range of opinion. Magnitude scaling of attitudes has been calibrated through numeric estimation. The following is an
example of numeric estimation of social opinion:2
I would like to ask your opinion about how serious you think certain crimes are. The first situation is, “A person steals a bicycle parked on the street.” This has been given a score of 10 to show its seriousness. Use this
situation to judge all others. For example, if you think a situation is 20 times more serious than the bicycle
theft, the number you tell me should be around 200, or if you think it is half as serious, the number you tell
me should be around 5, and so on.
COMPARED
TO
THE
BICYCLE
THEFT
AT
SCORE
10,
how
serious
is:
A parent beats his young child with his fists. The child requires hospitalization. A person plants a bomb
in a public building. The bomb explodes and 20 people are killed. . . .
Magnitude scaling of this type has shown some interesting results, but there are problems with
the technique. The researcher must be sure that the respondents have the competence to make these
proportional judgments, which means that respondents must be allowed to practice before attempting the actual research questions.
264
TYPES
OF
ATTITUDE RATING SCALES
TYPES OF ATTITUDE RATING SCALES
There are many ways to present a respondent with a continuum of numbered categories that
represent the range of possible attitude judgments. Figure 11-1 shows one of the ways the various attitudinal scales used in marketing research can be classified. They can be generally classified as single-item scales and multiple-item scales.
Single-Item Scales
As the name itself suggests, single-item scales are those that have only one item to measure a
construct. Under the single-item scales, the itemized-category scale is the most widely used by
marketing researchers. In some situations, comparative scales, rank-order scales, or constant-sum
scales have advantages. Each of these major types of rating scales will be discussed in turn.
Itemized-Category Scales. The following scale from the HMO study (discussed in Chapter 3) is
an itemized-category scale. There are four categories from which respondents can choose to indicate their overall level of satisfaction with their present health insurance plan:
_____
_____
_____
_____
Very satisfied
Quite satisfied
Somewhat satisfied
Not at all satisfied
This satisfaction scale has the following characteristics:
1. All categories are labeled.
2. The respondent is forced to make a choice; there is no provision for neutral opinion or “don’t
know” responses.
3. There are more favorable than unfavorable categories, so the scale is unbalanced.
4. There is no explicit comparison of the respondent’s present plan with other health insurance
plans.
Attitude Scales
Single-Item
Scales
ItemizedCategory
Scales
Comparative
Scales
Rank-Order
Scales
ConstantSum
Scales
Multi-Item
Scales
PairedComparison
Scales
Q-sort
Scales
Pictorial
Scales
Semantic
Differential
Scales
Likert
Scales
Associative
Scales
Thurstone
Scales
Stapel
Scales
FIGURE 11-1
Classification of attitude scales.
265
CHAPTER 11 ATTITUDE MEASUREMENT
1. Extent of category description
All categories labeled
Polar categories labeled
2. Treatment of respondent uncertainty or ignorance
Forced choice (no neutral point)
Neutral point
Provision of “don’t know” category
3. Balance of favorable and unfavorable categories
Equal
Unbalanced
4. Comparison judgment required
Yes
No
FIGURE 11-2
Types of itemized category scales.
The design of the satisfaction scale requires decisions along several dimensions, as shown in
Figure 11-2. Note that for this scale there is no attempt to make the intervals between categories
even approximately the same so this is an ordinal scale.
A quite different scale would have resulted had the decision been to label only the polar or
end categories, balance the favorable and unfavorable categories, and provide an obvious neutral category.
A numerical scaling of the response categories from 2 to 2, as presented in the following
chart, could help to create a scale that can be treated as an interval scale. This would be referred
to as a quasi-interval scale. While evidence is mixed, it appears that an adequate approximation
of an interval scale can be achieved with this procedure.3
Very
Satisfied
2
Very
Dissatisfied
1
0
1
2
Yet another set of choices is illustrated by the following scale, which was used to ask respondents in the HMO study about the various medical services (including private clinics, private doctors, and an existing HMO), on a number of attributes, one of which was quality of care provided.
Q28 . . . “we are interested in your opinions about the medical services offered in this area . . .
(a) Quality of medical care provided: (check one for each provider)”
266
Excellent
Very
Good
Average
Below
Average
Don’t
Know
Private doctors in area
Private clinics
Health organizations
in the area
The scale used for this question is unbalanced, with all categories labeled and a “don’t know”
category provided, and implies a comparison with other health-care providers. The decision to
use an unbalanced scale was based on an assumption of positively skewed attitudes toward all
health-care providers. This assumption was borne out by the results, which showed that only 15
percent rated private doctors as average or below average. However, 31 percent replied “don’t
know” to this question, indicating that they did not have sufficient experience to form a judgment. This could be thought of as a form of awareness question. In general, a “don’t know”
category should be provided whenever respondents may have insufficient experience to make a
meaningful attitude judgment.
TYPES
OF
ATTITUDE RATING SCALES
Comparative Scales. Another version of the preceding scale would label the categories “excellent,”
“very good,” “good,” “fair,” and “poor,” thereby eliminating the implicit comparison. The problem with a comparative scale is that the reference point is unclear and different respondents may
use different reference points or standards. Are private doctors rated “excellent” or “very good”
because they are superior to the existing alternatives, or because they measure up to an ideal form
of medical care provider? In marketing studies where competitive alternatives are being evaluated,
some form of explicit or implicit comparison should be built into the scale; for example:
Compared to private clinics in the area, the doctors in private
practice provide a quality of medical care which is
Very
superior
_____
_____
Neither superior
nor inferior
_____
_____
_____
Very
inferior
_____
_____
A recent review of research on the question of the appropriate number of response categories
concluded:4
■ ”Scales with two or three response alternatives generally are inadequate in that they are incapable of transmitting very much information and they tend to frustrate and stifle respondents.”
■ There is little to be gained from using more than nine categories.
■ An odd rather than an even number of categories is preferable when the respondent legitimately can adopt a neutral position.
Rank-Order Scales. Rank-order scales require the respondent to arrange a set of objects with
regard to a common criterion: advertisements in terms of interest, product features in terms of
importance, or new-product concepts with regard to willingness to buy in the future. The result is
an ordinal scale with the inherent limitations of weak scale properties. Ranking is widely used in
surveys, however, because it corresponds to the choice process occurring in a shopping environment where a buyer makes direct comparisons among competing alternatives (brands, flavors, product variations, and so on). An example of a rank-order scale is given below.
Please rank from 1 to 6 the following characteristics of the cellular phone service (1 is most important and 6
is least important, no ties allowed)
Characteristics
Rank
Total cost of service
______
Reception clarity
______
Low fixed cost of service
______
Reliability of service
______
24-hour customer service
______
Size of local coverage area
______
Rank-order scales are not without problems. Ranking scales are more difficult than rating
scales because they involve comparisons, and hence require more attention and mental effort.
The ranking technique may force respondents to make choices they might not otherwise make,
which raises the issue of whether the researcher is measuring a real relationship or one that is
artificially contrived.
Due to the difficulties of rating, respondents usually cannot meaningfully rank more than
five or six objects. The problem is not with the rankings of the first and last objects but with
267
CHAPTER 11 ATTITUDE MEASUREMENT
those in the undifferentiated middle. When there are several objects, one solution is to break the
ranking task into two stages. With nine objects, for example, the first stage would be to rank the
objects into classes: top three, middle three, and bottom three. The next stage would be to rank
the three objects within each class.
When using paired comparisons, the objects to be ranked are presented two at a time, and
the respondent has to choose between them according to some criterion such as overall preference or willingness to buy. Before a ranking of all objects can be obtained, all possible combinations of pairs have to be presented. This means that for n objects there are [n(n 1)兾2]
comparisons. This is very manageable for five objects (10 comparisons), but with more objects
the task can get out of hand. When 10 brands, for example, there are 45 paired comparisons.
Paired-comparisons data have some potential analytical advantages, which will become apparent in Chapter 22, when we discuss multidimensional scaling. A serious problem, however, is
that the comparison of two objects at a time is seldom the way choices are made in the marketplace; thus, an item may do well in a paired-comparison situation but perform poorly
in an actual market situation.5 Despite these limitations, however, rankings still have much to
recommend them if a researcher is interested in how consumers rank alternatives.
Q-Sort Scaling. When the number of objects or characteristics that are to be rated or ranked is very
large, it becomes rather tedious for the respondent to rank order or do a pairwise comparison. If the
respondent is forced to do a rank ordering or a pairwise comparison, a number of problems and
biases creep into the study. To deal with such a situation, the Q-sort scaling process is used. In
Q-sort scaling the respondents are asked to sort the various characteristics or objects that are being
compared into various groups, such that the objects in each group are viewed as the most similar.
Then the respondent can rank order the cards within each group which would be a much easier task.
For example, let us take the case of a toy manufacturing company such as Toys ‘R’ Us developing a
new product. After a marathon brainstorming session, the new-product team has come up with a
hundred different products, each with minor variations in features, and wants to test and find out
from consumers which feature combination is the most preferred and will generate the maximum
sales. The best scaling procedure that can be used in this context is Q-sort scaling. The procedure the
respondent is asked to use is as follows.
Each respondent is handed 100 cards, each containing a product with various features. The respondent is then
asked to sort the cards into 12 different piles in such a way that one pile contains what they feel is the most preferred among the products that have been developed, and another pile contains the least preferred of the products
that have been developed. The other 10 piles will contain cards with products that vary gradually from those with
higher preference to those with lower preference. The number of cards in each pile is normally distributed as shown
in Figure 11-3. In this particular case, only five cards can be placed in the most and the least preferred product
piles. After placing all the cards in the piles, the respondent is asked to rank-order only those products in the mostpreferred pile or in the top few sets of piles.
In Q-sort scaling, a relatively large number of groups or piles should be used (10 or more).
This increases the reliability or precision of the results.
Constant-Sum Scales. Constant-sum scales require respondents to allocate a fixed number of rating points (usually 100) among several objects, to reflect the relative preference for each object.6
It is widely used to measure the relative importance of attributes, as in the following example.
Please divide 100 points among the following characteristics so the division reflects the relative importance
of each characteristic to you in the selection of a health care plan.
268
Ability to choose a doctor
Extent of coverage provided
Quality of medical care
Monthly cost of the plan
Distance to clinic or doctor from your home
TOTAL
_____
_____
_____
_____
_____
100
TYPES
OF
ATTITUDE RATING SCALES
12
Number of cards in pile
11
10
9
8
7
6
5
1
2
3
4
5
6
7
Pile number
8
9
10
11
FIGURE 11-3
Plot of number of cards in each pile.
The most attractive feature of this scale is the quasi-interval nature of the resulting scale.
However, just how close it comes to a true interval scale has not been fully established. The scale
is limited in the number of objects or attributes it can address at one time. Respondents sometimes have difficulty allocating points accurately among more than a few categories.
Pictorial Scales. In pictorial scales the various categories of the scale are depicted pictorially. The
respondents are shown a concept or read an attitudinal statement and are asked to indicate their
degree of agreement or interest by indicating the corresponding position on the pictorial scale.
Therefore, in designing a format, it is of prime importance to design one that the respondent will
comprehend and that will enable him or her to respond accurately. Commonly used pictorial
scales are the thermometer scale and the funny faces scale. Pictorial scales are used mainly when
the respondents are young children or with people who are illiterate.
Paired-Comparison Scales. The brands to be rated were presented two at a time, so each brand
in the category was compared once to every other brand. In each pair the respondents were
asked to divide 10 points among the brands, on the basis of how much they liked one compared
to the other. A score was then totaled for each brand. Although this scale performs well on the
criteria, it is cumbersome to administer. Another possible limitation is that the frame of reference is always the other brands in the set being tested. These brands may change over time.
Several features are shared by all three of the most effective scales:
■ They restrict the numbers of highly positive ratings that can be given, either by forcing a
choice or by comparing brands directly.
■ They provide a limited number of categories that have verbal anchors. Respondents prefer
words to numbers, and especially avoid negative numbers. Also, including more than seven
categories may actually reduce the scale’s power to discriminate.
■ The stimulus to the respondent is simple and unambiguous. One of the worst-performing
scales presented a picture of a thermometer with 10 categories of liking; each was labeled by
a number from 0 to 100, as well as an assortment of verbal anchors. For example, 80 was
labeled “like very much” while 50 was “indifferent” and 30 was “not so good.”
269
CHAPTER 11 ATTITUDE MEASUREMENT
Although these issues are useful, and should be carefully considered, the best guidance
still comes from carefully tailoring the scale to the research objectives, followed by thorough pretesting for comprehension and discrimination.
Issues in Designing Single-Item Scales. Attitude rating scales are widely used to test the effectiveness of advertising copy or compare the performance of new product concepts and segment
markets. Despite years of experience with these applications, the design of the rating scale is usually an ad hoc judgment based on the researcher’s preferences and past experiences in similar situations. The various decisions that a researcher has to make regarding the form and structure of the
scale while designing a scale are described briefly below:
1. Number of scale categories. Theoretically, the number of rating-scale categories can vary from two
to infinity. A continuous rating scale has infinite categories, whereas the number of categories
in a discontinuous scale depends on several factors, such as the capabilities of the scalers, the
format of the interview, and the nature of the object.7 For example, if the survey is done by
telephone, the number of categories that a scale can have is very limited, because the memory
of the respondent is limited.
2. Types of poles used in the scale. All rating scales have verbal descriptors or adjectives that serve
as end points or anchors. The scale can have a single pole or two poles. An example of a
two-pole scale is “sweet . . . not sweet,” and an example of a scale with a single pole is the
Stapel scale, which is discussed later. The advantage of the single-pole scale over the scale
with double poles is ease of construction, as one need not look for adjectives to achieve bipolarity. The disadvantage is that we do not know what each category represents in a singlepole scale.
3. Strength of the anchors. By strength of the anchor we refer to the intensity of the adjective that
is used to anchor the scale. A rating-scale anchor could vary from “extremely colorful” to
“very colorful” to “colorful.” Anchor strength has been found to shape scale-response distributions; the stronger the anchors, the less likely scalers are to use the extreme scale categories,
so the resulting scale response distribution will be more peaked.
4. Labeling of the categories. Another decision that has to be made while developing scales is
whether to label every category of the scale or to label only the extreme categories. Labeling
all categories reduces the scale’s ambiguity.8 Evidence also shows that using such terms as
“very” or “somewhat” markedly influences responses to scales.9
5. Balance of the scale. A related decision is whether category labels should be balanced or unbalanced. A balanced four-category scale to measure the smell of a perfume could be
The smell of Morning Dew is…
_____ Very good
_____ Good
_____ Bad
_____ Very bad
while a corresponding unbalanced scale might be expressed as
_____ Superb
_____ Very good _____ Good
_____ Average
Generally, a balanced scale is preferred to an unbalanced scale in order to obtain meaningful results.
270
There is little argument on the criteria a rating scale ideally should satisfy. The results should
be reliable and valid, and there should be a sharp discrimination among the objects being rated
and a sensitivity to advertising or product stimuli. These criteria are seldom employed in practice. Part of the reason is the sheer variety of rating scales. The real problem is the absence of
empirical evidence on the performance of the various rating scales on these criteria. However,
one study of different scales did shed some useful light on the subject, and can help us narrow
down the set of acceptable scales.10
TYPES
OF
ATTITUDE RATING SCALES
Respondents in the study were given various subsets of the scales, and asked to rate six brands
in each of six package-goods categories such as coffee, analgesics, detergents, and toothpaste. Three
criteria were used to compare the performance of the scales: (1) response distribution, which is the
ability to avoid having responses pile up in the end categories; (2) discrimination among brands in
the category; and (3) concurrent validity—how well the ratings related to current brand usage.
Two scales were found to be particularly attractive:
■ Brand Awareness Scale. This question asked: When I mention detergents, what brand do you
think of? Any others? Have you heard of (interviewer mentions other brands of interest that
were not reported)?
_____ First unaided mention
_____ Second unaided mention
_____ Other unaided mention
_____ Aided recall
_____ Never heard of
This scale was consistently the best discriminator among brands and had high concurrent
validity. By design, it yielded uniform distributions of responses.
■ Verbal Purchase Intent Scale. The question asked: What is the chance of your buying (brand)
the next time you purchase this product?
Definitely
buy
_____
Probably
buy
_____
Might
buy
_____
Probably
not buy
_____
Definitely
not buy
_____
This balanced scale made efficient use of the five categories, distributing the responses quite
uniformly. The labels were easy for the respondents to handle. On average it discriminated
well.
Multiple Item Scales
Attitudes toward complex objects such as health plans, automobiles, credit instruments, or transportation modes have many facets. Thus it is often unrealistic to attempt to capture the full
picture with one overall attitude-scale question. For example, the public appears to support the
general idea of income tax reform but opposes the elimination of the most popular tax loopholes.
While beliefs in any specific issue, aspect, or characteristic are useful indicators of the overall
attitude, there may be unusual reasons that make the single belief unrepresentative of the general position.11 To cope with this problem, a variety of methods have been developed to measure a sample of beliefs toward the attitude objects (such as agreement or disagreement with a
number of statements about the attitude object) and combine the set of answers into some form
of average score. The most frequently employed of these methods are Likert, Thurstone, and
semantic-differential scales. An adaptation of these methods, with particular relevance to marketing problems, is associative scaling.
Likert Scales. Likert scales require a respondent to indicate a degree of agreement or disagreement with a variety of statements related to the attitude or object. They are also called
summated scales, because the scores on the individual items are summed to produce a total
score for the respondent. A Likert scale usually consists of two parts, the item part and the
evaluative part. The item part is essentially a statement about a certain product, event, or attitude. The evaluative part is a list of response categories ranging from “strongly agree” to
“strongly disagree.” An important assumption of this scaling method is that each of the items
(statements) measures some aspect of a single common factor; otherwise, the items cannot legitimately be summed. In other words, the resulting scale is unidimensional. The Likert scaling
method, then, refers to the several steps in the procedure for culling out the items that do not
271
CHAPTER 11 ATTITUDE MEASUREMENT
belong. The result is a series of 5 to 20 or more statements and questions, of which those given
below are illustrative.
Agree
Strongly
Agree
Somewhat
Neither
Agree nor
Disagree
Disagree
Somewhat
Disagree
Strongly
1. There needs to be
much improvement in
the health insurance
available for people
like me.
2. I have a variety of
very good health
plans from which to
choose.
3. I haven’t heard of a
health insurance plan
that will protect me
against a disastrous
illness.
Thurstone Scales. The procedure of Thurstone scales is also known as the method of equalappearing intervals, since the objective is to obtain a unidimensional scale with interval
properties.
The first step is to generate a large number of statements or adjectives reflecting all degrees of
favorableness toward the attitude objects. Then, a group of judges is given this set of items (as many
as 75 to 100 in all) and asked to classify them according to their degree of favorableness or unfavorableness. Usually, this is done with an 11-category bipolar scale, with “very favorable” at one
end, “very unfavorable” at the other, and a neutral position in the middle. The judges are instructed
to treat the intervals between categories as equal and to make evaluations of each item without
expressing their own attitudes. The scale value of each item is the median position to which it is
assigned by the judges. Items that have been placed in many different categories are discarded as
ambiguous because there was no consensus among the judges. The resulting scale consists of 10 to
20 items that are distributed uniformly along the scale of favorability. The scale then is administered
as part of a survey by asking each respondent to select those statements which best reflect his or
her feelings toward the attitude object. The respondent’s attitude score is the average of the scale
scores of the chosen statements.
Because of the two-stage procedure, a Thurstone scale is both time consuming and expensive to construct; however, the scale itself is easy to administer and requires a minimum of
instructions. Because there is not an explicit response to each item in the scale, it does not
have as much diagnostic value as a Likert scale. Thurstone scales also have been criticized
because the scale values themselves may depend on the attitudes of the original judges. This
seems to be a problem only with topics that elicit strong feelings, such as abortion or school
integration.
Semantic-Differential Scales. Semantic-differential scales are used widely to describe the set
of beliefs that comprise a person’s image of an organization or brand. The procedure is also an
insightful procedure for comparing the images of competing brands, stores, or services.12 Respondents are asked to rate each attitude object in turn on a number of five- or seven-point rating
scales, bounded at each end by polar adjectives or phrases. Some researchers prefer unipolar
scales, while others use bipolar scales. In either case, the respondent chooses the end point only
if that adjective is closely descriptive of that object. However, the midpoint of the scale has two
272
TYPES
OF
ATTITUDE RATING SCALES
different meanings, depending on the type of scale. With unipolar scales, the midpoint is
simply a step on the scale from “sweet” to “not sweet,” whereas on a bipolar scale it is a neutral point.
Low price
______
______
_______
______
______
High price
Consistent quality
______
______
_______
______
______
Spotty quality
Tangy
______
______
_______
______
______
Smooth
Bitter
______
______
_______
______
______
Not bitter
There may be as many as 15 to 25 semantic-differential scales for each attitude object. The
scales in the preceding chart were used in a beer-brand image study in a U.S. regional market.
(Only 4 of a total of 10 scales are shown.) Each of 10 brands was evaluated separately on the same
set of 10 scales, for comparison.
This set of scales is characteristic of most marketing applications of the semantic differential:
1. The pairs of objects or phrases are selected carefully to be meaningful in the market being
studied and often correspond to product or service attributes.13 Exploratory research generally is required to ensure that important attributes are represented and described in words
that are familiar to respondents.14
2. The negative or unfavorable pole is sometimes on the right side and sometimes on the left.
This rotation is necessary to avoid the halo effect, in which the location of previous judgments
on the scale affects subsequent judgments because of respondent carelessness.
3. The category increments are treated as interval scales, so group mean values can be computed
for each object on each scale. As with Likert scaling, this assumption is controversial, but is
adopted because it permits more powerful methods of analysis to be used.
The semantic differential also may be analyzed as a summated rating scale. Each of the seven
scale categories is assigned a value from 3 to 3 or 1 to 7, and the scores across all adjective
pairs are summed for each respondent. Individuals then can be compared on the basis of their
total scores. Summation is not usually advisable, however, for a good deal of specific information is lost in the aggregate score, which may be distorted if there are several scales that measure roughly the same thing.
Profile Analysis. Profile analysis is an application of the semantic-differential scale. Visual
comparisons of the images of different objects can be aided by plotting the mean ratings for
each object on each scale. To show what can be done, Figure 11-4 compares the ratings for
two well-known national brands of beer and a regional brand, on 6 of the 10 scales. Even
with three brands and only 6 of 10 attributes, the interpretation of the profiles is not easy.
With more brands and attributes, the overall comparisons of brands are even harder to grasp.
A second difficulty is that not all attributes are independent; that is, several of the attributes
may be measuring approximately the same dimension. For example, to most beer drinkers
there is not likely to be much difference in the meaning of the “tangy–smooth” and
“bitter–not bitter” scales. This is borne out by the similarity of the scores of the three brands
on these two scales in Figure 11-4. Fortunately, there are several procedures using multidimensional scaling techniques that can deal effectively with these problems and yield easily
interpreted spatial maps that describe the overall image of a brand. These are discussed in
detail in Chapters 21 and 22, which deal with methods for analyzing interdependencies
among variables.
Stapel Scales. Stapel scales are simplified versions of semantic-differential scales, which use
only one pole rather than two. Respondents are asked to indicate the object by selecting a numerical response category. The higher the positive score, the better the adjective describes the object.
273
CHAPTER 11 ATTITUDE MEASUREMENT
Heavy
Light
Regional
brand
Low price
High price
Millers
Spotty quality
Consistent quality
Budweiser
Tangy
Smooth
Poor flavor
Good flavor
Bitter
1
2
3
4
5
Not bitter
Mean Rating
FIGURE 11-4
Profile analysis of three beer brands.
A typical format for this scale is shown for Coors beer as it would be adapted to the measurement of beer brand attitudes.
The main virtue of this scale is that it is easy to administer and construct, because there is
no need to provide adjectives or phrases to assure bipolarity.15
Coors Beer
3
3
2
Heavy
3
12
2
1
Consistent
11
1
quality
1
1
2
2
2
3
3
3
Tangy
1
Associative Scaling. Although the semantic-differential and Stapel scales are used widely for
image studies, they have substantial limitations in markets where the average respondent is
likely to be knowledgeable only about a small subset among a large number of choice alternatives. They also can be cumbersome and time consuming to administer when there are a number of attributes and alternatives to consider. An alternative approach, associative scaling,
designed to overcome these limitations, asks the respondent simply to associate one alternative
with each question. The questions in Figure 11-5 illustrate how this approach is employed in a
telephone survey of retail-store images.
The technique is argued to be particularly appropriate to choice situations that involve a
sequential decision process. For example, supermarkets have to be within a reasonable distance
to be considered, and within that set the choice is made on the basis of which chain is best in
satisfying customers’ needs. Of course, the technique does not answer the questions of how consumers make trade-offs when there are several important dimensions and no alternative is superior across the board. Thus the benefits of low cost and ease of telephone administration are
purchased at a possible cost of reduced validity in representing the market structure. For this
reason the associative technique is best suited to market tracking, where the emphasis is on
understanding shifts in relative competitive positions.
274
Eaton’s
Store
or
Catalog
The
Bay
Sayvette
Towers
Woolco
Zellers
Any
Other
None
DK
1
2
3
4
5
6
7
8
9
0
B
X
2. Has the highest overall
prices?
1
2
3
4
5
6
7
8
9
0
B
X
3. Is the easiest one to get to
from your home?
1
2
3
4
5
6
7
8
9
0
B
X
4. Has the most
knowledgeable, helpful
sales clerks?
1
2
3
4
5
6
7
8
9
0
B
X
5. Has the highest-quality
products?
1
2
3
4
5
6
7
8
9
0
B
X
6. Has the lowest-quality
products?
1
2
3
4
5
6
7
8
9
0
B
X
7. Gives you the best overall
value for the money?
1
2
3
4
5
6
7
8
9
0
B
X
8. Gives you the worst
overall value for the
money?
1
2
3
4
5
6
7
8
9
0
B
X
9. Has the best advertising?
1
2
3
4
5
6
7
8
9
0
B
X
10. Is the best for the latest,
most fashionable
merchandise?
1
2
3
4
5
6
7
8
9
0
B
X
11. Has the largest overall
merchandise selection
or assortment?
1
2
3
4
5
6
7
8
9
0
B
X
12. Do you shop at most
often?
1
2
3
4
5
6
7
8
9
0
B
X
Which store
1. Has the lowest overall
prices?
275
ATTITUDE RATING SCALES
FIGURE 11-5
Retail-store image questions (telephone questionnaire).
OF
Horizon
More
Than
One
Answer
TYPES
Simpsons
Sears
Store
or
Catalog
CHAPTER 11 ATTITUDE MEASUREMENT
GENERAL GUIDELINES FOR DEVELOPING
A MULTIPLE-ITEM SCALE
Multiple-item scales very often are used in social sciences research to measure abstract constructs. The characteristic that is to be measured is generally referred to as the construct. Most
of the well-known scales that measure constructs, such as IQ, consumer confidence, and so on,
are multiple-item scales. Developing a multiple-item scale is a complex procedure and requires
quite a lot of technical expertise. Figure 11-6 presents the various steps in the development of
a multiple-item scale. The stages of development are discussed below.
1. Determine clearly what it is that you want to measure. The scale should be well grounded in theory. Relevant social science theories should always be considered before developing the scale.
The construct to be measured and the scale itself should be specific. The meaning and the definition of the construct should be clearly distinguishable from other constructs.
2. Generate as many items as possible. Items essentially are statements that are relevant to the
construct. The content of each item should reflect primarily the construct of interest. If items
are being written anew, they should be written with as much creativity as possible. The
greater the number of initial items generated, the better the final scale will be. The items
that are developed should not be too long, nor should they pose any reading difficulty to
the respondent.
Determine clearly what you
are going to measure.
Generate as many
items as possible.
Ask experts in the field
to evaluate the initial
pool of items.
Determine the type of
attitudinal scale
to be used.
Include some items that
will help in the
validation of the scale.
Administer the items to
an initial sample.
Evaluate and refine
the items.
Finally, optimize the
scale length.
276
FIGURE 11-6
Steps in multiple-item scale development.
INTERPRETING ATTITUDE SCALES
3. Ask experts to evaluate the initial pool of items. Experts are people who have worked or are
currently working on the phenomenon that is being studied. They can be either business
managers or academics who are doing research on that particular phenomenon. Having
experts review the item pool can confirm or invalidate the definition of the construct.
Experts can also give inputs on the relevancy, clarity, and conciseness of the items. Based
on the experts’ evaluation, the initial pool of items is modified. Some items are dropped
while others are added, and a few of the existing ones are changed.
4. Determine the type of attitudinal scale to be used. The next step in the multiple-item scaling
process is to decide on the type of scale to be employed. The various scales such as the Likert scale, semantic-differential scale, Thurstone, and associative scales have already been discussed. In fact, the type of attitudinal scale to be used has to be decided quite early, because
the wording of the items varies with each scale format.
5. Include validation items in the scale. Certain items are added to the scale in order to improve
the scale’s validity and also to detect certain flaws in it. An example of items that can be
included in the scale to increase the validity of the results is those that are socially desirable.
Some respondents answer in a certain fashion because they want to be perceived as socially
desirable. In order to weed out such answers, items of social desirability are added to the
scale. Those responses that correlate highly with social desirability are then dropped.
6. Administer the items to an initial sample. Once it has been determined which construct-related
items are to be included in the scale, the next step is to administer the scale to an initial sample. This is done to check the validity of the items. This will yield the best results if the sample size is large and the sample is representative of the population.
7. Evaluate and refine the items of the scale. The ultimate quality that is sought in an item is high
correlation with the true score of the latent variable that is being measured. The properties that the items of a scale should possess are high intercorrelation, high item scale correlation, high item variances, a mean close to the center of the range of possible scores,
and a high coefficient alpha. The items in the scale are then evaluated on the basis of these
criteria.
8. Optimize scale length. The larger the scale, the greater the reliability, but shorter scales are easier for the respondent to answer. Hence, a balance has to be struck between brevity and reliability, and the optimal scale length has to be determined. Certain items in the scale are
dropped or modified, then the final scale is ready to be administered to the respondent.
INTERPRETING ATTITUDE SCALES
Conclusions obtained from attitude-scale measurements are strictly limited by the properties of
the scale that is used. Failure to recognize these limits can lead to serious misinterpretation, as
we see from Marketing Research in Action 11-1. The problem was created by assuming a ratio
scale where there was really only an interval scale.
MARKETING RESEARCH IN ACTION 11-1
Interpreting Attitude Scales: A Problem for the
Advertising Review Board
The Phoenix Drug Co. currently sells the leading brand of tranquilizers, known as Restease. A competitor, Montfort Drug Co., recently announced a tranquilizer brand called Calm, which they claim to
be 50 percent more effective in reducing tension than the leading brand.
As product manager for Restease you are concerned and also angry, because you don’t believe there
is a significant difference in effectiveness. Your first action is to complain to the National Advertising
Review Board (NARB), an advertising industry-sponsored body that investigates advertising claims and
can put considerable pressure on advertisers to change their claims.
277
CHAPTER 11 ATTITUDE MEASUREMENT
MARKETING RESEARCH IN ACTION 11-1 (continued)
As part of the investigation of your complaint, the research director from Montfort Drug is asked by
the NARB to present the research findings that support the claim. Among the findings are the results of
an apparently well-designed comparison test with large sample sizes. In the test, one group of product
users was given Restease capsules. After a month, each user was asked to rate the effectiveness of the
brand as follows:
For easing tension I found Restease to be:
Very
effective
Effective
Neither
effective nor
ineffective
Ineffective
Very
ineffective
Another group of product users, identical in all respects to the first group, was given Calm capsules and
asked to rate the effectiveness of this brand on the same scale.
The research director for Montfort Drug coded the scale with a 2 for “very effective,” 1 for “effective,” 0 for “neither effective nor ineffective,” 1 for “ineffective,” and 2 for “very ineffective.” The
director correctly points out that this is a well-accepted coding convention. When the data for the two
groups are summarized, the average response for the Calm user groups is calculated to be 1.2, while
the average for Restease is 0.8. Because the 0.4 difference is 50 percent more than the 0.8 level
achieved by Restease, the director concludes that the claims of superior effectiveness are valid.
While you are listening to this argument, the research director from your company has taken the same
data and calculated that Calm is only 10.5 percent more effective, rather than 50 percent as claimed.
Immediately you examine the figures. The only difference is that the “very effective” category has been
coded 1, and “very ineffective” assigned 5, with the middle category assigned 3. It is argued that
this is an equally acceptable coding procedure. The two different coding schemes are as follows:
Very
effective
2
1
Effective
1
2
Neither
effective nor
ineffective
0
3
Ineffective
1
4
Very
ineffective
2
5
Soon you will be asked to present the basis of your complaint to the Review Board. What do you say
about the capacity of the data presented by your competitor to support its claim of superiority?
SOURCE: Adapted from B. Venkatesh, “Unthinking Data Interpretation Can Destroy Value of Research,” Marketing News, January 27, 1978, pp. 6, 9.
CHOOSING AN ATTITUDINAL SCALE
The choice of an appropriate scale is complicated by two problems:
1. There are many different techniques, each with its own strengths and weaknesses.
2. Virtually any technique can be adapted to the measurement of any one of the attitude components.
278
While these problems are significant impediments to broad generalizations, it is also true that all
techniques are not equally suitable for all purposes. Table 11-2 summarizes some useful rules of
thumb as to which scale types are likely to be best suited to the various components of attitudes.
What is most evident from this table is the versatility of the itemized-category scale, which itself
has many variations. Ultimately, the researcher’s choice will be shaped by (1) the specific information that is required to satisfy the research objectives, (2) the adaptability of the scale to the
data collection method and budget constraints, and (3) the compatibility of the scale with the
structure of the respondent’s attitude.
CHOOSING
AN
ATTITUDINAL SCALE
TABLE 11-2
Appropriate Applications of Various Attitude Scales
Type of Scale
Itemized
Category
Rank
Order
Constant Sum
Likert
Knowledge
awareness
attribute beliefs
attribute importance
a
a
a
b
b
b
a
b
b
a
Affect or liking
overall preferences
specific attributes
a
a
b
b
a
b
b
b
b
a
Action
intentions
a
b
a
b
Attitude Component
Semantic
Differential
a Very appropriate
b Sometimes appropriate
The value of careful selection and adaptation of scales will be demonstrated by a study of
attitudes toward automobile dealers. The data came from a mail survey of people who had purchased a new car from an automobile dealer between one and two years earlier.16 Each respondent rated 14 attributes of the dealer’s service department, on two separate scales. The first scale
asked how important the attribute was (on a four-point scale ranging from “extremely important” to “not important”). The second scale asked how well the service department performed (on
a four-point scale from “excellent” to “poor”). Each attribute was located in the grid shown in
Figure 11-7, according to its median score on the two scales.
A. Concentrate here
B. Keep up the good work
Extremely
1 important
2
3
4
5
6
7
8
9
10
Fair
performance
Excellent
performance
11
12
13
14
C. Low priority
Slightly
important
FIGURE 11-7
Importance–performance grid.
D. Possible overkill
279
CHAPTER 11 ATTITUDE MEASUREMENT
TABLE 11-3
Comparison of Two Rating Techniques
Attribute
Median
Importance
Rating
Illustrative
Constant-Sum Scale
Customer A
Customer B
1. Job done right the first time
3.8
35
25
2. Fast action on complaints
3.6
15
20
3. Prompt warranty work
3.6
10
15
4. Able to do any job needed
3.6
10
10
5. Service available when needed
3.4
15
20
6. Courteous and friendly service
3.4
15
10
100
100
To illustrate how the grid can be used, quadrant A includes those attributes, such as “low
prices on service” (attribute 10), which are very important, and where performance is rated only
fair. By contrast, in quadrant B, customers place a high value on “courteous and friendly service”
(attribute 6) and “prompt warranty work” (attribute 3), and are pleased with the dealer’s performance. In quadrant C, the dealer’s performance is rated low in terms of providing courtesy
buses and rental cars, but fortunately these are not perceived as important services.
While the importance–performance grid yields useful insights, this application suffered from
the inability of the four-point rating scale to discriminate. Ten of the 14 attributes had median
ratings between 3.3 and 3.8 on